{ "complete": true, "report": { "bucket": "vn_S_ASMR_reg", "label": "Strong ASMR Style", "definition": { "id": "vn_S_ASMR_reg", "category": "talking_style", "head": "vn_S_ASMR_reg", "pole": "high", "label": "Strong ASMR Style", "floor": 4.0, "canonical_dimension_name": "ASMR Style" }, "recipe_status": "direct_manual_recipe", "manual_provenance": [ { "id": "rank4/style_s_asmr/full-v2-sealed/sealed-selected", "mission": "style_s_asmr", "format_status": "CANONICAL", "heldout": true, "qualified": true, "source_cfg": 3, "source_model": { "label": "Humaneness 1.5 + real-speech rank 4", "repo": "laion/Humaneness-Voice-Large-1.5-RealSpeech-LoRAs", "revision": "fed33e554f094fcdc89fc0124aed2f9d42018714", "rank": 4 }, "metrics": { "ce": 6.127020049095154, "genuineness": 4.496991658210755, "wer": 0.008846538070328393, "identity": 0.5479652185738086, "voiceclap": 0.07065998954698444, "fitness": 0.7365491728619838, "steering_v2": 0.805349731337989, "raw_target": 2.8389798223972322, "target_cdf": 0.805349731337989, "family_hit_all": null, "nominal_hit_all": null, "n": 100, "valid_fraction": 1, "by_arm": { "reference": { "n": 50, "ce": 6.184008722305298, "genuineness": 4.528309106826782, "wer": 0.00589247311827957, "identity": 0.5479652185738086, "fitness": 0.7333224111252634, "steering_v2": 0.7884799382492357 }, "none": { "n": 50, "ce": 6.0700313758850095, "genuineness": 4.465674209594726, "wer": 0.011800603022377216, "identity": null, "fitness": 0.7397759345987042, "steering_v2": 0.8222195244267423 } } } } ], "selected_wording": { "id": "manual-763a50e7d58f", "general": "Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.", "inline": [ "Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.", "Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.", "Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable." ], "cfg": 3, "recipe": "P06", "strategy": "initial-physical-directing", "timing_syntax": "canonical", "canonical_labels": true, "neutral_event": "keep", "parent": "cfg-3", "origin": "exact_frozen_manual_recipe" }, "selection": { "manual_no_adapter": { "config": { "scale": 0.0, "cfg": 4.0 }, "n": 20, "target_z": 1.5871093689205047, "target_raw": 3.785932159423828, "ce": 5.791680431365966, "genuineness": 4.4166583061218265, "identity": 0.5327619522809982, "wer": 0.033089955280357755, "valid_fraction": 1, "identity_delta": -0.010125356912612914, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, "manual_adapter": { "config": { "scale": 0.5, "cfg": 5.0 }, "n": 20, "target_z": 1.6792968510873185, "target_raw": 3.913032603263855, "ce": 5.655265045166016, "genuineness": 4.366136193275452, "identity": 0.5062399044632911, "wer": 0.06332752837977296, "valid_fraction": 0.9, "identity_delta": -0.03664740473031998, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, "grid": [ { "config": { "scale": 0.0, "cfg": 4.0 }, "n": 20, "target_z": 1.5871093689205047, "target_raw": 3.785932159423828, "ce": 5.791680431365966, "genuineness": 4.4166583061218265, "identity": 0.5327619522809982, "wer": 0.033089955280357755, "valid_fraction": 1, "identity_delta": -0.010125356912612914, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 3.0 }, "n": 20, "target_z": 1.4933593278501365, "target_raw": 3.6566773891448974, "ce": 5.994019961357116, "genuineness": 4.491657018661499, "identity": 0.5958077058196067, "wer": 0.0029411764705882353, "valid_fraction": 1, "identity_delta": 0.05292039662599564, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 2.5 }, "n": 20, "target_z": 1.2162109455384438, "target_raw": 3.2745682179927824, "ce": 6.129839658737183, "genuineness": 4.437996840476989, "identity": 0.5945256501436234, "wer": 0.006458333333333333, "valid_fraction": 1, "identity_delta": 0.051638340950012206, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 2.0 }, "n": 20, "target_z": 1.0975586173587066, "target_raw": 3.1109802424907684, "ce": 6.134208154678345, "genuineness": 4.4869499683380125, "identity": 0.5761884853243828, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.03330117613077164, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 1.5 }, "n": 20, "target_z": 0.92451171223143, "target_raw": 2.872397541999817, "ce": 6.18464412689209, "genuineness": 4.474084138870239, "identity": 0.6139326646924019, "wer": 0.0059649122807017545, "valid_fraction": 1, "identity_delta": 0.07104535549879074, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 1.25 }, "n": 20, "target_z": 0.5366210897983281, "target_raw": 2.337606167793274, "ce": 6.20728075504303, "genuineness": 4.45494225025177, "identity": 0.6069724321365356, "wer": 0.006274509803921568, "valid_fraction": 1, "identity_delta": 0.0640851229429245, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 1.0 }, "n": 20, "target_z": 0.4019531035191167, "target_raw": 2.151937133073807, "ce": 6.2267402648925785, "genuineness": 4.449293708801269, "identity": 0.5645257391035556, "wer": 0.0059649122807017545, "valid_fraction": 1, "identity_delta": 0.021638429909944534, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 5.0 }, "n": 20, "target_z": 1.6078124746012197, "target_raw": 3.814475882053375, "ce": 5.646230292320252, "genuineness": 4.3626843929290775, "identity": 0.4876794621348381, "wer": 0.028198529411764706, "valid_fraction": 1, "identity_delta": -0.05520784705877304, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 5.0 }, "n": 20, "target_z": 1.6792968510873185, "target_raw": 3.913032603263855, "ce": 5.655265045166016, "genuineness": 4.366136193275452, "identity": 0.5062399044632911, "wer": 0.06332752837977296, "valid_fraction": 0.9, "identity_delta": -0.03664740473031998, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 4.0 }, "n": 20, "target_z": 1.6234374929748054, "target_raw": 3.8360183596611024, "ce": 5.778177952766418, "genuineness": 4.402223396301269, "identity": 0.4960890904068947, "wer": 0.02389877880976952, "valid_fraction": 1, "identity_delta": -0.04679821878671646, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 4.0 }, "n": 20, "target_z": 1.6054687486490011, "target_raw": 3.811244547367096, "ce": 5.771426653861999, "genuineness": 4.405047643184662, "identity": 0.5041621148586273, "wer": 0.042976651186790504, "valid_fraction": 0.95, "identity_delta": -0.03872519433498382, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 4.0 }, "n": 20, "target_z": 1.5980468575721136, "target_raw": 3.801011860370636, "ce": 5.735486030578613, "genuineness": 4.377433109283447, "identity": 0.558050774037838, "wer": 0.015465686274509803, "valid_fraction": 1, "identity_delta": 0.015163464844226838, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 3.0 }, "n": 20, "target_z": 1.5343749989192008, "target_raw": 3.713226389884949, "ce": 5.969397687911988, "genuineness": 4.486949992179871, "identity": 0.5546415612101555, "wer": 0.009607843137254901, "valid_fraction": 1, "identity_delta": 0.011754252016544342, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 3.0 }, "n": 20, "target_z": 1.4593749989192009, "target_raw": 3.609822618961334, "ce": 5.936634182929993, "genuineness": 4.413520312309265, "identity": 0.6002537414431572, "wer": 0.006458333333333333, "valid_fraction": 1, "identity_delta": 0.05736643224954605, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 3.0 }, "n": 20, "target_z": 1.440039052299958, "target_raw": 3.583163821697235, "ce": 5.952718067169189, "genuineness": 4.428896594047546, "identity": 0.5529676869511604, "wer": 0.012916666666666667, "valid_fraction": 1, "identity_delta": 0.010080377757549285, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 2.5 }, "n": 20, "target_z": 1.3078125005403995, "target_raw": 3.400860834121704, "ce": 6.000771260261535, "genuineness": 4.454314565658569, "identity": 0.5835059344768524, "wer": 0.009399509803921569, "valid_fraction": 1, "identity_delta": 0.04061862528324127, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 2.5 }, "n": 20, "target_z": 1.2178711006063196, "target_raw": 3.2768571019172668, "ce": 6.027081370353699, "genuineness": 4.39437837600708, "identity": 0.6249458998441696, "wer": 0.006666666666666666, "valid_fraction": 1, "identity_delta": 0.08205859065055847, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 2.5 }, "n": 20, "target_z": 1.1964843678397055, "target_raw": 3.2473708510398867, "ce": 5.9942186117172245, "genuineness": 4.419796323776245, "identity": 0.5427191436290741, "wer": 0.006666666666666666, "valid_fraction": 1, "identity_delta": -0.00016816556453704833, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 2.0 }, "n": 20, "target_z": 1.1214843764860989, "target_raw": 3.143967092037201, "ce": 6.079205107688904, "genuineness": 4.46404242515564, "identity": 0.5525043770670891, "wer": 0.02245614035087719, "valid_fraction": 0.95, "identity_delta": 0.009617067873477936, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 2.0 }, "n": 20, "target_z": 1.114550777500978, "target_raw": 3.1344076216220857, "ce": 6.109983015060425, "genuineness": 4.449293744564057, "identity": 0.573352724313736, "wer": 0.005409356725146199, "valid_fraction": 1, "identity_delta": 0.030465415120124816, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 2.0 }, "n": 20, "target_z": 1.0306640533807572, "target_raw": 3.0187515735626222, "ce": 6.105813121795654, "genuineness": 4.452117943763733, "identity": 0.5878884389996528, "wer": 0.006666666666666666, "valid_fraction": 1, "identity_delta": 0.04500112980604172, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 1.5 }, "n": 20, "target_z": 0.7907226707394638, "target_raw": 2.6879403233528136, "ce": 6.171538710594177, "genuineness": 4.469377076625824, "identity": 0.5933203071355819, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.05043299794197083, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 1.5 }, "n": 20, "target_z": 0.7860351669566668, "target_raw": 2.6814775824546815, "ce": 6.194175362586975, "genuineness": 4.455255913734436, "identity": 0.5668412245810032, "wer": 0.006458333333333333, "valid_fraction": 1, "identity_delta": 0.023953915387392045, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 1.25 }, "n": 20, "target_z": 0.7224121215171031, "target_raw": 2.59375941157341, "ce": 6.184246969223023, "genuineness": 4.474397945404053, "identity": 0.6136667564511299, "wer": 0.013859649122807016, "valid_fraction": 0.95, "identity_delta": 0.07077944725751877, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 1.25 }, "n": 20, "target_z": 0.7042480270659779, "target_raw": 2.5687162667512893, "ce": 6.193381071090698, "genuineness": 4.479104948043823, "identity": 0.5508964359760284, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.008009126782417298, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 1.5 }, "n": 20, "target_z": 0.6402832005074396, "target_raw": 2.480526876449585, "ce": 6.190203976631165, "genuineness": 4.415403091907502, "identity": 0.5881660848855972, "wer": 0.0029411764705882353, "valid_fraction": 1, "identity_delta": 0.04527877569198609, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 1.25 }, "n": 20, "target_z": 0.6088378936140851, "target_raw": 2.437172698974609, "ce": 6.233888697624207, "genuineness": 4.468435668945313, "identity": 0.590385428071022, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.047498118877410886, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 1.0 }, "n": 20, "target_z": 0.4357910014563985, "target_raw": 2.1985900163650514, "ce": 6.210457777976989, "genuineness": 4.446469533443451, "identity": 0.5435333665460348, "wer": 0.005263157894736842, "valid_fraction": 1, "identity_delta": 0.000646057352423668, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 1.0 }, "n": 20, "target_z": 0.38364257158954257, "target_raw": 2.1266920924186707, "ce": 6.187424063682556, "genuineness": 4.447097134590149, "identity": 0.5797351963818074, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.036847887188196184, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 1.0 }, "n": 20, "target_z": 0.3171386819230547, "target_raw": 2.035002052783966, "ce": 6.247391223907471, "genuineness": 4.386219501495361, "identity": 0.5619056962430478, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.01901838704943657, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 5.0 }, "n": 20, "target_z": 1.6835937367602103, "target_raw": 3.918956792354584, "ce": 5.6086018800735475, "genuineness": 4.366136193275452, "identity": 0.46579391658306124, "wer": 0.03053921568627451, "valid_fraction": 0.95, "identity_delta": -0.07709339261054993, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 5.0 }, "n": 20, "target_z": 1.599218755133796, "target_raw": 3.8026275753974916, "ce": 5.6548679113388065, "genuineness": 4.349818587303162, "identity": 0.4910649709403515, "wer": 0.05320519435844513, "valid_fraction": 0.9, "identity_delta": -0.05182233825325966, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" } ], "development_standard": [ { "scale": 0.0, "n": 20, "target_z": 0.024951178140257636, "target_raw": 1.6321581900119781, "ce": 6.176304292678833, "genuineness": 4.34824948310852, "identity": 0.5428873091936112, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.0, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "scale": 0.5, "n": 20, "target_z": 0.2147460995255205, "target_raw": 1.8938316643238067, "ce": 6.2176062107086185, "genuineness": 4.411637568473816, "identity": 0.5662717670202255, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.02338445782661438, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "scale": 0.75, "n": 20, "target_z": 0.31035155932515246, "target_raw": 2.0256445318460465, "ce": 6.207280707359314, "genuineness": 4.354839301109314, "identity": 0.5597197011113166, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.016832391917705535, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "scale": 1.0, "n": 20, "target_z": 0.17988282148414303, "target_raw": 1.845765072107315, "ce": 6.189012598991394, "genuineness": 4.388416111469269, "identity": 0.5859634682536126, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.043076159060001375, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" } ] }, "heldout": { "standard": { "n": 50, "target_z": -0.03717772730589798, "target_raw": 1.546500015258789, "ce": 6.277096824645996, "genuineness": 4.413583059310913, "identity": 0.5892024129629135, "wer": 0.007492941840767927, "valid_fraction": 1, "identity_delta": 0.0, "identity_pairs": 50, "gates_pass": true, "gate_scope": "heldout_50_reference_pairs" }, "manual": { "n": 50, "target_z": 1.3238281180018254, "target_raw": 3.422941837310791, "ce": 5.897397422790528, "genuineness": 4.447975788116455, "identity": 0.5337682750821113, "wer": 0.01115050472029877, "valid_fraction": 1, "identity_delta": -0.055434137880802155, "identity_pairs": 50, "gates_pass": false, "gate_scope": "heldout_50_reference_pairs" }, "adapter": { "n": 50, "target_z": 1.3501562403268474, "target_raw": 3.4592408657073976, "ce": 5.709553251266479, "genuineness": 4.376679944992065, "identity": 0.4986606428027153, "wer": 0.04616327875023527, "valid_fraction": 0.94, "identity_delta": -0.09054177016019821, "identity_pairs": 50, "gates_pass": false, "gate_scope": "heldout_50_reference_pairs" }, "matched_no_adapter": { "n": 50, "target_z": 1.3375000060524753, "target_raw": 3.441791501045227, "ce": 5.7461292171478275, "genuineness": 4.370278406143188, "identity": 0.521460379064083, "wer": 0.0458408868019852, "valid_fraction": 0.96, "identity_delta": -0.06774203389883042, "identity_pairs": 50, "gates_pass": false, "gate_scope": "heldout_50_reference_pairs" } }, "adapter_vs_matched_cfg": { "n": 50, "delta_z": 0.012656234274372201, "ci95_z": [ -0.0596875228048623, 0.09015622303406093 ], "language_deltas": { "en": -0.04648442836445862, "de": -0.09765625675499474, "fr": 0.13984372162902203, "nl": 0.07773437864769714, "es": -0.010156243785404806 }, "languages_improved": 2, "clear_gain": false, "delta_ce": -0.036575965881347657, "delta_genuineness": 0.006401538848876953, "delta_wer": 0.00032239194825007216, "note": "Exploratory language-cluster bootstrap; five voices, one fixed reference voice per language. No familywise claim across72targets." }, "adapter_vs_best_manual": { "n": 50, "delta_z": 0.026328122325022088, "ci95_z": [ -0.019296868380105144, 0.0690624957848833 ], "language_deltas": { "en": -0.012499987030410043, "de": 0.07421875513379607, "fr": 0.0484374583892323, "nl": -0.0527343700013039, "es": 0.07421875513379601 }, "languages_improved": 3, "clear_gain": false, "delta_ce": -0.18784417152404786, "delta_genuineness": -0.07129584312438965, "delta_wer": 0.0350127740299365, "note": "Exploratory language-cluster bootstrap; five voices, one fixed reference voice per language. No familywise claim across72targets." }, "manual_vs_standard": { "n": 50, "delta_z": 1.3610058453077234, "ci95_z": [ 1.1811425629431493, 1.583095682025774 ], "language_deltas": { "en": 1.1839355200407833, "de": 1.1122070169631861, "fr": 1.3851562437854048, "nl": 1.7821288821008938, "es": 1.3416015636483491 }, "languages_improved": 5, "clear_gain": true, "delta_ce": -0.37969940185546874, "delta_genuineness": 0.03439272880554199, "delta_wer": 0.0036575628795308427, "note": "Exploratory language-cluster bootstrap; five voices, one fixed reference voice per language. No familywise claim across72targets." }, "route": { "adapter": false, "scale": 0.0, "cfg": 1.0, "reason": "No new configuration passes the held-out gates; retain standard fallback and inspect audio", "fallback_standard": true }, "reference_only": true, "updated": 1793025026.2673802 }, "rows": [ { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "000dbc7bfa3272004c97658865721773803c9bb1993cea15363ef0a23ac6e684", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f690d36e4c7967cfe84727a69fd6d9c9cf081c724a55a337d0b09b474e1ecb7c", "generation_seconds_batch": 25.95690680795815, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7627739906311035, "target_z": 1.5703124702780231, "ce": 5.95112943649292, "genuineness": 4.400654315948486, "identity": 0.5580289959907532, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/000dbc7bfa3272004c97658865721773803c9bb1993cea15363ef0a23ac6e684.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "0098ba987f6672a1b2b3f4ea6d956fa61033d9f777be43aed2b77f51fa4dbfa5", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0cb2b43613bbff6cb988b8978999999ffc89fd6fa09c30012a8d49dcab054a53", "generation_seconds_batch": 23.912861598073505, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1255476474761963, "target_z": 0.3828124918940063, "ce": 6.149696350097656, "genuineness": 4.256305694580078, "identity": 0.25671443343162537, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het Oude Raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0098ba987f6672a1b2b3f4ea6d956fa61033d9f777be43aed2b77f51fa4dbfa5.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "00ac3034e38e98729d9a04507f34254c442d2bf8ac0ba62dc80a440883ce9993", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "821a0bb72feb0bb158a312b3428398da8ce1dfc4c1c9f8be982f345dde88e2ad", "generation_seconds_batch": 17.477941902936436, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3211536407470703, "target_z": 1.249999913536067, "ce": 6.149696350097656, "genuineness": 4.65797233581543, "identity": 0.48125195503234863, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/00ac3034e38e98729d9a04507f34254c442d2bf8ac0ba62dc80a440883ce9993.mp3" }, { "id": "heldout-es-3", "language": "es", "sentences": [ "Alguien ha movido la pequeña silla azul.", "Estaba aquí... junto a la estantería alta.", "La colocaremos después de comer." ], "text": "Alguien ha movido la pequeña silla azul. Estaba aquí... junto a la estantería alta. La colocaremos después de comer.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-3", "key": "011d08ee1879187e3cf2c82c9cd13bb3a1cdfba8cb9f2b6ad107413117674a49", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La colocaremos después de comer.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Natural clear speech.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Natural clear speech.) [3.2 seconds duration] La colocaremos después de comer.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0a1625d02674deb811410433f09d1af9819623ab5ac6ce233717326bd35f175c", "generation_seconds_batch": 22.70087477308698, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.278069019317627, "target_z": 1.2187501188879077, "ce": 5.5539960861206055, "genuineness": 4.65797233581543, "identity": 0.4570211172103882, "hypothesis": "Alguien ha movido la pequeña silla azul. Estaba aquí, junto a la estantería alta. La colocaremos después de comer.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/011d08ee1879187e3cf2c82c9cd13bb3a1cdfba8cb9f2b6ad107413117674a49.mp3" }, { "id": "heldout-fr-2", "language": "fr", "sentences": [ "La route tourne à gauche après le pont.", "Continue... jusqu’à la petite barrière en bois.", "Je t’attendrai de l’autre côté." ], "text": "La route tourne à gauche après le pont. Continue... jusqu’à la petite barrière en bois. Je t’attendrai de l’autre côté.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-2", "key": "016239452bce0ce4332878a92270f6b34dae323bdee878a035dd2e253d358502", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Natural clear speech.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Natural clear speech.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1f6afa526479171af0e9ef5c2d2143744f6bde65f21f8df4ecf2a6d1e17ecf46", "generation_seconds_batch": 24.339199988986365, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7304604053497314, "target_z": 1.5468750378279708, "ce": 5.720792293548584, "genuineness": 4.545003414154053, "identity": 0.7224752306938171, "hypothesis": "La route tourne à gauche après le pont, continue jusqu'à la petite barrière en bois. Je t'attendrai de l'autre côté.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/016239452bce0ce4332878a92270f6b34dae323bdee878a035dd2e253d358502.mp3" }, { "id": "heldout-nl-0", "language": "nl", "sentences": [ "Je bericht kwam nadat ik vertrokken was.", "Ik las het twee keer... en legde mijn telefoon weg.", "Zeg maar wanneer we kunnen praten." ], "text": "Je bericht kwam nadat ik vertrokken was. Ik las het twee keer... en legde mijn telefoon weg. Zeg maar wanneer we kunnen praten.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-0", "key": "0170747f91cbbc9fab78357d4775472d9b98729a9c2e430bb222675cc5ddcd98", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Natural clear speech.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "09e4bec0e5d3048cc0f0227d29d9ad8fafdfe735b0006eaa97435d43d0712aec", "generation_seconds_batch": 15.034878571983427, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0949580669403076, "target_z": 1.0859375405299685, "ce": 5.7684478759765625, "genuineness": 4.256305694580078, "identity": 0.4779410660266876, "hypothesis": "Je bericht kwam nadat ik vertrokken was. Ik las het twee keer en legde mijn telefoon weg. Zeg maar wanneer we kunnen praten.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0170747f91cbbc9fab78357d4775472d9b98729a9c2e430bb222675cc5ddcd98.mp3" }, { "id": "heldout-fr-4", "language": "fr", "sentences": [ "J’ai écrit le numéro sur l’enveloppe.", "Lis-le bien... et vérifie les deux derniers chiffres.", "Appelle-moi quand tu auras un moment." ], "text": "J’ai écrit le numéro sur l’enveloppe. Lis-le bien... et vérifie les deux derniers chiffres. Appelle-moi quand tu auras un moment.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-4", "key": "01f5d5f22d9b51f4e674eec92be80d0710e62f5704093e3409bc811236eec3e5", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1fa4112c3680032019e185e22ce6b5c6d944a2ab25a73e858e14c95e5a02416a", "generation_seconds_batch": 17.850444269017316, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5385158061981201, "target_z": -0.04296877026498429, "ce": 6.046441555023193, "genuineness": 4.545003414154053, "identity": 0.8097183108329773, "hypothesis": "J'ai écrit le numéro sur l'enveloppe, lis le bien et vérifie les deux derniers chiffres. Appelle-moi quand tu auras un moment.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/01f5d5f22d9b51f4e674eec92be80d0710e62f5704093e3409bc811236eec3e5.mp3" }, { "id": "heldout-fr-3", "language": "fr", "sentences": [ "Quelqu’un a déplacé la petite chaise bleue.", "Elle était ici... près de la grande bibliothèque.", "Nous la remettrons après le déjeuner." ], "text": "Quelqu’un a déplacé la petite chaise bleue. Elle était ici... près de la grande bibliothèque. Nous la remettrons après le déjeuner.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-3", "key": "027e077f7a176865f9620685615b30acbf4c3939f901789b9a818d829b37842a", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Natural clear speech.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Natural clear speech.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5d09a6c4c503771f2af752a27643772b4a5ced6b87792e111fa58df0fc6e6eec", "generation_seconds_batch": 15.301845128997229, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.752002716064453, "target_z": 1.5624999351520503, "ce": 5.681078910827637, "genuineness": 4.482243061065674, "identity": 0.7955862283706665, "hypothesis": "Quelqu'un a déplacé la petite chaise bleue. Elle était ici, près de la grande bibliothèque. Nous la remettrons après le déjeuner.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/027e077f7a176865f9620685615b30acbf4c3939f901789b9a818d829b37842a.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "028899e31b379eb06673b5bbe74f71e9dd455191a4bc88ddb26aa8a15f8ac9f7", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e1d8561055304a2f09c1e32d3abb74c05a2090fb1e29be50c14b0dccc59abbc6", "generation_seconds_batch": 23.671042102971114, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0841867923736572, "target_z": 1.078125005403996, "ce": 6.268836498260498, "genuineness": 4.425758361816406, "identity": 0.5465285778045654, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/028899e31b379eb06673b5bbe74f71e9dd455191a4bc88ddb26aa8a15f8ac9f7.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "02d6f916af4682351523db48f7a86c8e88a289349e801582e96a33bb5986b1a6", "bucket": "vn_S_ASMR_reg", "phase": "wording", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8d65ca395b6999a605605d1c102527e268cc2563d4c611843e9ec4c2c1478ee0", "generation_seconds_batch": 17.731394647969864, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8903045654296875, "target_z": 0.9374998919200838, "ce": 6.157639026641846, "genuineness": 4.582659721374512, "identity": 0.7559173107147217, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/02d6f916af4682351523db48f7a86c8e88a289349e801582e96a33bb5986b1a6.mp3" }, { "id": "heldout-de-2", "language": "de", "sentences": [ "Die Straße führt hinter der Brücke nach links.", "Geh weiter... bis zu dem kleinen Holztor.", "Ich warte auf der anderen Seite." ], "text": "Die Straße führt hinter der Brücke nach links. Geh weiter... bis zu dem kleinen Holztor. Ich warte auf der anderen Seite.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-2", "key": "02ea6fdaba2773b70ee6d21d6706eb4527d17e6bcd170090ee1b8c2e3deef4b0", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Natural clear speech.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Natural clear speech.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "bf6fa5cb4c017f03cb738d6e22f75f3eb8c6f99102738be34c6a45e8cc4aea1b", "generation_seconds_batch": 21.965058194007725, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5581207275390625, "target_z": 1.421874994596004, "ce": 5.649308204650879, "genuineness": 4.319066047668457, "identity": 0.5092787742614746, "hypothesis": "Die Straße führt hinter der Brücke nach links. Geh weiter bis zu dem kleinen Holztor. Ich warte auf der anderen Seite.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/02ea6fdaba2773b70ee6d21d6706eb4527d17e6bcd170090ee1b8c2e3deef4b0.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "0316aeaf6b30e1469fe4da0a09827c26fbd6e50144b4b61191a37652af1d443d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "39905aedba0a9b31a0d7ab884aef79094a50f4ed10bc2eb13d09e7bd08bff730", "generation_seconds_batch": 23.29884066595696, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6442904472351074, "target_z": 1.4843749297480544, "ce": 5.982900142669678, "genuineness": 4.394378662109375, "identity": 0.510712742805481, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0316aeaf6b30e1469fe4da0a09827c26fbd6e50144b4b61191a37652af1d443d.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "0407fe1f5871f285f614730ba55d799047946be7a3679d12f323e99e7c49c860", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "eda84a6f522ad424b3c8ee37057b162159acd5d535e5f7781ad7f045b8ac9c7f", "generation_seconds_batch": 24.273448113002814, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.734121799468994, "target_z": 0.8242186513770765, "ce": 6.4912309646606445, "genuineness": 4.444586753845215, "identity": 0.31331032514572144, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0407fe1f5871f285f614730ba55d799047946be7a3679d12f323e99e7c49c860.mp3" }, { "id": "heldout-fr-3", "language": "fr", "sentences": [ "Quelqu’un a déplacé la petite chaise bleue.", "Elle était ici... près de la grande bibliothèque.", "Nous la remettrons après le déjeuner." ], "text": "Quelqu’un a déplacé la petite chaise bleue. Elle était ici... près de la grande bibliothèque. Nous la remettrons après le déjeuner.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-3", "key": "0422f00a8e9351a3ffb415d0aa40fbb15d9c2fcf6f039b63a7e4a24f9641d908", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Natural clear speech.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Natural clear speech.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "38f9c19ae345a50659d64a14aceece64714e55e07f3a5c623771c28a2cb05c5f", "generation_seconds_batch": 15.300432551070116, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5796632766723633, "target_z": 1.4375000648479497, "ce": 5.502368927001953, "genuineness": 4.507347106933594, "identity": 0.716942548751831, "hypothesis": "Quelqu'un a déplacé la petite chaise bleue. Elle était ici, près de la grande bibliothèque. Nous la remettrons après le déjeuner.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0422f00a8e9351a3ffb415d0aa40fbb15d9c2fcf6f039b63a7e4a24f9641d908.mp3" }, { "id": "heldout-de-3", "language": "de", "sentences": [ "Jemand hat den kleinen blauen Stuhl umgestellt.", "Er stand hier... neben dem hohen Bücherregal.", "Wir stellen ihn nach dem Essen zurück." ], "text": "Jemand hat den kleinen blauen Stuhl umgestellt. Er stand hier... neben dem hohen Bücherregal. Wir stellen ihn nach dem Essen zurück.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-3", "key": "043810c5844198d0e8597d485c00d5b43526bf88d92f9d2ae2c028c1108a39bc", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Natural clear speech.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Natural clear speech.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f252dea03db1ac6f5f6d01e7938b78d6767d0ce955f2d171f0fdf6be2fb4a730", "generation_seconds_batch": 17.473335723974742, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4934933185577393, "target_z": 1.3749999567680335, "ce": 5.808161735534668, "genuineness": 4.545003414154053, "identity": 0.38849708437919617, "hypothesis": "Jemand hat den kleinen blauen Stuhl umgestellt. Er stand hier, neben dem hohen Bücherregal. Wir stellen ihn nach dem Essen zurück.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/043810c5844198d0e8597d485c00d5b43526bf88d92f9d2ae2c028c1108a39bc.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "04592797ae55b332e707cc7d9cd081f1b947aa8ba764fe78258fa65f8d978c60", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "978db225836f7be2b26c5845c2301d662dfc16b25d3343247097da68805b1174", "generation_seconds_batch": 25.36465780204162, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8428030014038086, "target_z": 0.17773439594048376, "ce": 6.324435234069824, "genuineness": 4.250029563903809, "identity": 0.3524519205093384, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/04592797ae55b332e707cc7d9cd081f1b947aa8ba764fe78258fa65f8d978c60.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "04bbbc8da974256ba0e75e40f55f629cac77e3f7a3737fa17906ac4e07e9d0d2", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "260993ea77738cb2502444d2093f954623cc77983cd9d2feb489a419233ddbf2", "generation_seconds_batch": 22.932823816896416, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.171325445175171, "target_z": 0.41601567971545755, "ce": 6.229123115539551, "genuineness": 4.507347106933594, "identity": 0.7440457344055176, "hypothesis": "ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/04bbbc8da974256ba0e75e40f55f629cac77e3f7a3737fa17906ac4e07e9d0d2.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "04c0d3d0feec531977c8af9454649bb6e00a0d39678bd72a0b04ef42cd4debe0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "061119eaf5585306379273a01d9594a4b55fce9440b083da455c3d6319bf4558", "generation_seconds_batch": 16.92559107101988, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.773545265197754, "target_z": 1.578125005403996, "ce": 6.197352409362793, "genuineness": 4.645420074462891, "identity": 0.8067313432693481, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/04c0d3d0feec531977c8af9454649bb6e00a0d39678bd72a0b04ef42cd4debe0.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "0538baa476d715876d83d79b8067e75434e6fab1fd24e75d792127380033abe3", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2e4c85b65f6fab0f11992eb2f691d1a763831b37cc7d6f11e19ef76f6b4e6cda", "generation_seconds_batch": 23.255606304039247, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8095204830169678, "target_z": 0.87890622433102, "ce": 6.157639026641846, "genuineness": 4.582659721374512, "identity": 0.7199484705924988, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0538baa476d715876d83d79b8067e75434e6fab1fd24e75d792127380033abe3.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "053c3dbf7743b9ca67d9acaaf23d2fc0dd907dd853048b2830dc43c5dc97ada0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 131, "seconds_generated": 10.48, "mp3_sha256": "6b461667b27e530fb4c85c41dc88b39115f4ea6e82127466bf488f6acefc9cb2", "generation_seconds_batch": 18.00180448696483, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6442904472351074, "target_z": 1.4843749297480544, "ce": 6.006728172302246, "genuineness": 4.331617832183838, "identity": 0.35576140880584717, "hypothesis": "I found the little box. It was beside the old window. We can look inside.", "wer": 0.0625, "duration_error_s": 0.28000000000000114, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/053c3dbf7743b9ca67d9acaaf23d2fc0dd907dd853048b2830dc43c5dc97ada0.mp3" }, { "id": "heldout-de-0", "language": "de", "sentences": [ "Deine Nachricht kam, als ich schon weg war.", "Ich las sie zweimal... dann legte ich das Handy weg.", "Sag mir, wann wir reden können." ], "text": "Deine Nachricht kam, als ich schon weg war. Ich las sie zweimal... dann legte ich das Handy weg. Sag mir, wann wir reden können.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-0", "key": "059867605942d1f4c46c2c44d936e08b7fbe89dd4feef4c05500a7ed426c27a3", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wann wir reden können.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wann wir reden können.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4a1fb932239a5e92dfee870aefb9cce1dde71e6ffd78713d0275ba76bdea5696", "generation_seconds_batch": 18.830065532005392, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1524758338928223, "target_z": 0.40234382970893817, "ce": 6.340320587158203, "genuineness": 4.557555198669434, "identity": 0.6676643490791321, "hypothesis": "Deine Nachricht kam, als ich schon weg war. Ich las sie zweimal, dann legte ich das Handy weg. Sag mir, wann wir reden können.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/059867605942d1f4c46c2c44d936e08b7fbe89dd4feef4c05500a7ed426c27a3.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "05abb45a4270315c116293ef4d3c5e8fed1f44850b3fa7863634eb947fe9a479", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 126, "seconds_generated": 10.08, "mp3_sha256": "c101234efd968da2cb00f19c4590fce7b75eb7c971eb6cfba2627bb45788bf76", "generation_seconds_batch": 27.22625131707173, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.2259368896484375, "target_z": 1.9062500972719245, "ce": 5.518253803253174, "genuineness": 4.582659721374512, "identity": 0.40216824412345886, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar junto.", "wer": 0.29411764705882354, "duration_error_s": 0.11999999999999922, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/05abb45a4270315c116293ef4d3c5e8fed1f44850b3fa7863634eb947fe9a479.mp3" }, { "id": "heldout-en-2", "language": "en", "sentences": [ "The road turns left behind the bridge.", "Keep going... until you reach the small wooden gate.", "I will meet you on the other side." ], "text": "The road turns left behind the bridge. Keep going... until you reach the small wooden gate. I will meet you on the other side.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-2", "key": "05e386a731e8174f71ee6b1868fdebd6799e95ee1d98f936935543778283c172", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The road turns left behind the bridge.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I will meet you on the other side.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The road turns left behind the bridge.\n(Natural clear speech.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Natural clear speech.) [3.2 seconds duration] I will meet you on the other side.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "50c78375aba1c2a7c30d58f5c7fb7cce8e3881a0e9f3ee87aa6b61f2673d2fad", "generation_seconds_batch": 25.159405322978273, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4719510078430176, "target_z": 1.359375059443954, "ce": 6.133810997009277, "genuineness": 4.281409740447998, "identity": 0.18913903832435608, "hypothesis": "The road turns left behind the bridge. Keep going until you reach this small wooden gate. I will meet you on the other side.", "wer": 0.041666666666666664, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/05e386a731e8174f71ee6b1868fdebd6799e95ee1d98f936935543778283c172.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "069eac0c47bcc1af59090a64298f610ee62a86f070764b5a376f95917a7a4962", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ff123f27dc1cdc1b8571a6dc435f7fa762cd8f31c03053b585e5e8cab01a72b6", "generation_seconds_batch": 21.867876461008564, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.527531147003174, "target_z": 2.1250000432319665, "ce": 5.736677646636963, "genuineness": 4.532451152801514, "identity": 0.824891984462738, "hypothesis": "La porte est encore ouverte, j'ai poussé ton manteau sur la chaise, dis-moi quand tu es prêt.", "wer": 0.05555555555555555, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/069eac0c47bcc1af59090a64298f610ee62a86f070764b5a376f95917a7a4962.mp3" }, { "id": "heldout-es-2", "language": "es", "sentences": [ "La carretera gira a la izquierda tras el puente.", "Sigue... hasta la pequeña puerta de madera.", "Te esperaré al otro lado." ], "text": "La carretera gira a la izquierda tras el puente. Sigue... hasta la pequeña puerta de madera. Te esperaré al otro lado.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-2", "key": "06e73435c8c4c831b2fc06dbb1f579120f495587eef24ed3a282fff6c3ab195d", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Te esperaré al otro lado.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Natural clear speech.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Natural clear speech.) [3.2 seconds duration] Te esperaré al otro lado.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "cf7b075b801c700d3e72b267f24c0d90840e902efc13bf35006a2e42bac51e8f", "generation_seconds_batch": 14.928427975974046, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4180946350097656, "target_z": 1.32031238381409, "ce": 5.681078910827637, "genuineness": 4.557555198669434, "identity": 0.5105081796646118, "hypothesis": "La carretera gira a la izquierda tras el puente. Sigue hasta la pequeña puerta de madera. Te esperaré al otro lado.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/06e73435c8c4c831b2fc06dbb1f579120f495587eef24ed3a282fff6c3ab195d.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "0703fc4a33550374414be6f2d6abaf9efc30f259a58bb842164d18834b793831", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f8ddc2d6ed28f31b3befdf3cd041cae6906bb83731ed009fd2999352cf41f2e9", "generation_seconds_batch": 23.839920007972978, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3211536407470703, "target_z": 1.249999913536067, "ce": 5.998785495758057, "genuineness": 4.300237655639648, "identity": 0.6644624471664429, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0703fc4a33550374414be6f2d6abaf9efc30f259a58bb842164d18834b793831.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "07264d71e4a1fdd911a19f76707d51a59bba43b4760ab9d40480892cf58ab11b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6f4ea3c59f241178650de510c29a5e505c32feda253024f9ef63fd03c234c5e4", "generation_seconds_batch": 17.863600790966302, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5581207275390625, "target_z": 1.421874994596004, "ce": 6.260893821716309, "genuineness": 4.519899368286133, "identity": 0.7775623798370361, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/07264d71e4a1fdd911a19f76707d51a59bba43b4760ab9d40480892cf58ab11b.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "0770c2156a28e216a69913519b517cd3b451efaafa8a2d14119a3ce0511b6711", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c40bce70c2a1f5cdd3ae912c27a616700d2d6d7d5dd8741b5bdd9a29c9b54a8f", "generation_seconds_batch": 21.304264158941805, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.206331729888916, "target_z": 0.4414061594830702, "ce": 6.403861999511719, "genuineness": 4.632867813110352, "identity": 0.5784394145011902, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0770c2156a28e216a69913519b517cd3b451efaafa8a2d14119a3ce0511b6711.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "0804b0820f8dbe2a6b5fc087eaed1a7cf191093ec85b76849a901d0233b95882", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ebb500d822066b8f2f65c1bfa8cbb7d30fb4584b66045d8be133596605dd649c", "generation_seconds_batch": 23.4669335790677, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.859714984893799, "target_z": 1.640624940556046, "ce": 6.0305562019348145, "genuineness": 4.325342178344727, "identity": 0.5208949446678162, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0804b0820f8dbe2a6b5fc087eaed1a7cf191093ec85b76849a901d0233b95882.mp3" }, { "id": "heldout-en-0", "language": "en", "sentences": [ "Your message arrived after I had left.", "I read it twice... before putting the phone away.", "Let me know when we can talk." ], "text": "Your message arrived after I had left. I read it twice... before putting the phone away. Let me know when we can talk.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-0", "key": "0838a3a65c5368467a3c449df911f09495ac9e60a7d4d0f04efd42fe621871e8", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Your message arrived after I had left.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Let me know when we can talk.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Your message arrived after I had left.\n(Natural clear speech.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Natural clear speech.) [3.2 seconds duration] Let me know when we can talk.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "596d5953afb8ceb4782eb551970a8d25e28293275a2c4e053a7052c1cfac1d39", "generation_seconds_batch": 16.887940023094416, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.224212646484375, "target_z": 1.179687443258044, "ce": 6.300607204437256, "genuineness": 4.419482707977295, "identity": 0.4764743149280548, "hypothesis": "Your message arrived after I had left. I read it twice before putting the phone away. Let me know when we can talk.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0838a3a65c5368467a3c449df911f09495ac9e60a7d4d0f04efd42fe621871e8.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "089e29e2f0585b01b2cb205ebfdc6abfbc751b7f9ddff657feb56f953707407e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2492e38dbd1778e12e836150963f3b2a94e22663333ffae340da032b50c3defe", "generation_seconds_batch": 18.202271878020838, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0411019325256348, "target_z": 1.0468750378279708, "ce": 5.585766792297363, "genuineness": 4.18726921081543, "identity": 0.4340534210205078, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/089e29e2f0585b01b2cb205ebfdc6abfbc751b7f9ddff657feb56f953707407e.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "08eed1744a11429ef69fe134c62e33cab723d2e050722b1f631f90005fa70547", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "be7e14108dd864718d418bd26e2f82a513ad4a13c1e6cff3ea2cbb1714e52ba0", "generation_seconds_batch": 17.277718389057554, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5940959453582764, "target_z": 0.7226562135230283, "ce": 5.974957466125488, "genuineness": 4.174716949462891, "identity": 0.7076432704925537, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/08eed1744a11429ef69fe134c62e33cab723d2e050722b1f631f90005fa70547.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "09ed39e83ea202959fa68d143e91baa710c03f8aaff05be2902f5196029e4470", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "7ea6a9f48fdf948785ac4ee68e8ab01236b6c4132e7966de714ff61883dd2aaf", "generation_seconds_batch": 25.206191510078497, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.182851791381836, "target_z": 1.8749999567680335, "ce": 5.506340026855469, "genuineness": 4.519899368286133, "identity": 0.479272723197937, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que.", "wer": 0.11764705882352941, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/09ed39e83ea202959fa68d143e91baa710c03f8aaff05be2902f5196029e4470.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "09f8600de03e78eebbaa20ec8bc5b1d245c1eec1bdee0d38d83b8d464de47cd0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "999b78b4715907676509a5904c90d53d36f94f980dc07c87d905f10e6d746de4", "generation_seconds_batch": 25.090834756032564, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2234574556350708, "target_z": -0.2714843851324921, "ce": 6.078212261199951, "genuineness": 4.482243061065674, "identity": 0.7640883922576904, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd, zeg maar, wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/09f8600de03e78eebbaa20ec8bc5b1d245c1eec1bdee0d38d83b8d464de47cd0.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "0aa2c3d0a815d2f0031cc7575de8072d88c1720b6859025a7da9c470f1d967d1", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "784060cae35cd532e2d5e1f4028850df0b8bbab2ad1728341df05ef2743b386e", "generation_seconds_batch": 24.81006707798224, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5581207275390625, "target_z": 1.421874994596004, "ce": 6.070269584655762, "genuineness": 4.381826400756836, "identity": 0.5524311065673828, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0aa2c3d0a815d2f0031cc7575de8072d88c1720b6859025a7da9c470f1d967d1.mp3" }, { "id": "heldout-nl-3", "language": "nl", "sentences": [ "Iemand heeft het kleine blauwe stoeltje verplaatst.", "Het stond hier... naast de hoge boekenkast.", "We zetten het na de lunch terug." ], "text": "Iemand heeft het kleine blauwe stoeltje verplaatst. Het stond hier... naast de hoge boekenkast. We zetten het na de lunch terug.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-3", "key": "0b8832f8b550af4df785fb99c35143359e355381caece9b16558657db7c501eb", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We zetten het na de lunch terug.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Natural clear speech.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Natural clear speech.) [3.2 seconds duration] We zetten het na de lunch terug.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "ec86f09740862375b98b4f0da2a8d62c07bda6aabcf027fa4e50587affe3fc80", "generation_seconds_batch": 17.493517817929387, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0087881088256836, "target_z": 1.0234374324500524, "ce": 6.356205940246582, "genuineness": 4.645420074462891, "identity": 0.5822725296020508, "hypothesis": "Iemand heeft het kleine blauwe stoeltje verplaatst. Het stond hier, naast de hoge boekenkast. We zetten het na de lunch terug.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0b8832f8b550af4df785fb99c35143359e355381caece9b16558657db7c501eb.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "0ba4311caef9a01e4ead3ade5e8ec579270dbab737388bcd91f55859f5de8b48", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6e808de58b12e2112db8fad1879e4ec176b6f326c6425f8472ebbe6a98072138", "generation_seconds_batch": 16.851059868000448, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.245755195617676, "target_z": 1.1953125135099896, "ce": 6.41180419921875, "genuineness": 4.7458367347717285, "identity": 0.6058732271194458, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0ba4311caef9a01e4ead3ade5e8ec579270dbab737388bcd91f55859f5de8b48.mp3" }, { "id": "heldout-fr-2", "language": "fr", "sentences": [ "La route tourne à gauche après le pont.", "Continue... jusqu’à la petite barrière en bois.", "Je t’attendrai de l’autre côté." ], "text": "La route tourne à gauche après le pont. Continue... jusqu’à la petite barrière en bois. Je t’attendrai de l’autre côté.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-2", "key": "0ba60843ddd3c94f1283263803143c257c530b959be7426547dc271388b5d30a", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Natural clear speech.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Natural clear speech.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "090d1694b10b5895c94d952e462d954021db7e41156524fbb18ab1f168b9d8f4", "generation_seconds_batch": 15.536081769037992, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.655061721801758, "target_z": 1.4921874648740272, "ce": 5.5817952156066895, "genuineness": 4.469690799713135, "identity": 0.7563143372535706, "hypothesis": "La route tourne à gauche après le pont, continue jusqu'à la petite barrière en bois, je t'attendrai de l'autre côté.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0ba60843ddd3c94f1283263803143c257c530b959be7426547dc271388b5d30a.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "0ba7f89efd97eed2d5cfed5c32046dc945471d36a740f44b8eca897895e13315", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6d266da60989f186647b2bf0a5945d94bc3d8b33a7ccc7fbf8c0a68410767fc9", "generation_seconds_batch": 23.488602872006595, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1703567504882812, "target_z": 1.140625113483912, "ce": 6.1894097328186035, "genuineness": 4.356721878051758, "identity": 0.5956605672836304, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0ba7f89efd97eed2d5cfed5c32046dc945471d36a740f44b8eca897895e13315.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "0bbb30619ae75defd643ff65776cdf2c27ebc9ed350a533ecaf5f0b455d9e8ca", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "cfdf74f0efff8828cf079a481c6ef7438071d815b1e72ba7b73610669074ebc6", "generation_seconds_batch": 22.884885416948237, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.010512351989746, "target_z": 1.750000086463933, "ce": 5.998785495758057, "genuineness": 4.632867813110352, "identity": 0.8492209911346436, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0bbb30619ae75defd643ff65776cdf2c27ebc9ed350a533ecaf5f0b455d9e8ca.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "0c550e38139f527bcba43ad87e473f788a10e7c7d8866e11c14cc8321ce27b2b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d9ef257cabaf78a635dafc70b4f4204fa1a816f7200379c9f588ec186fd1bae2", "generation_seconds_batch": 17.110385830979794, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2099933624267578, "target_z": -0.2812500540399581, "ce": 6.109982967376709, "genuineness": 4.331617832183838, "identity": 0.5959113240242004, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0c550e38139f527bcba43ad87e473f788a10e7c7d8866e11c14cc8321ce27b2b.mp3" }, { "id": "heldout-nl-4", "language": "nl", "sentences": [ "Ik schreef het nummer op de envelop.", "Lees het goed... en controleer de laatste twee cijfers.", "Bel me wanneer je even tijd hebt." ], "text": "Ik schreef het nummer op de envelop. Lees het goed... en controleer de laatste twee cijfers. Bel me wanneer je even tijd hebt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-4", "key": "0cb70e86df8754d3bba44fe7f798aac7d0ffe2ca68fc75b92ff1a1ed29986ba7", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2e836b3e6c4b1740f6f6f001106bccd03ad73417fcfa107ef8b0c5bbb94a91a7", "generation_seconds_batch": 20.571306576952338, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6714321374893188, "target_z": -0.6718749945960042, "ce": 6.221180438995361, "genuineness": 4.381826400756836, "identity": 0.639276385307312, "hypothesis": "Ik schreef het nummer op de envelop. Lees het goed en controleer de laatste twee cijfers. Bel me wanneer je even tijd hebt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0cb70e86df8754d3bba44fe7f798aac7d0ffe2ca68fc75b92ff1a1ed29986ba7.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "0d1b4c1025e9212b8d1868626c4a64580ecbbf35a34535eb0ab3719163f8ed89", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6f90fe1e35721b1d10ec3af873a93050eafa15709a505e2325a01df5932d6ac9", "generation_seconds_batch": 24.797911969013512, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8428030014038086, "target_z": 0.17773439594048376, "ce": 6.213237762451172, "genuineness": 3.9676074981689453, "identity": 0.2387797236442566, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0d1b4c1025e9212b8d1868626c4a64580ecbbf35a34535eb0ab3719163f8ed89.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "0d5eaee13d78f7ae21b62b1b7137e4d369489b304895ce228f89175653a9f1f0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6abc7ffd0189b0f7824ef177c67a26ed6c6dd0b514587e9e282b9de8b7793622", "generation_seconds_batch": 21.867876461008564, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.505988597869873, "target_z": 2.109374972980021, "ce": 5.720792293548584, "genuineness": 4.570107460021973, "identity": 0.8374456763267517, "hypothesis": "La porte est encore ouverte, j'ai posé ton manteau sur la chaise, dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0d5eaee13d78f7ae21b62b1b7137e4d369489b304895ce228f89175653a9f1f0.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "0d752dfddc728bdd9fb121fd23d0b793e8b11a506eab3862a7e2055b282e232c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "71d791d46a5f20a09099e661d8ad9ce88f9c792f04b96cd09ec0a5a17d856d7a", "generation_seconds_batch": 16.090442794957198, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5456254482269287, "target_z": 0.6874999783840168, "ce": 6.403861999511719, "genuineness": 4.532451152801514, "identity": 0.5270771980285645, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0d752dfddc728bdd9fb121fd23d0b793e8b11a506eab3862a7e2055b282e232c.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "0d8f9441baa7f9e5eea1461e5424fd3406c196668fcc967da84afa6e13037fb8", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "89a09de212f1b9711a8bf691bea0552cb6616c13146006e58a0ec83193d1112d", "generation_seconds_batch": 16.5183989610523, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9135711193084717, "target_z": 1.679687443258044, "ce": 6.109982967376709, "genuineness": 4.65797233581543, "identity": 0.7784085273742676, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0d8f9441baa7f9e5eea1461e5424fd3406c196668fcc967da84afa6e13037fb8.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "0ddf3bf4383de9101d3c978ad23a630762e781ddbad09710b08e057b16f9fd11", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b9eb0a56d1a2094e6a0eb8952bc8cf95977812be78bf1ad94abf4d272e2c288a", "generation_seconds_batch": 17.2146433070302, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.376733779907227, "target_z": 2.01562489732408, "ce": 6.006728172302246, "genuineness": 4.507347106933594, "identity": 0.7431917786598206, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0ddf3bf4383de9101d3c978ad23a630762e781ddbad09710b08e057b16f9fd11.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "0df0c727931b4bcdba28b9929a1d67e74c449c017e3b3f076b2f48f636d75bbb", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "59c26a15a5cab7dd033850b42a3ba4f07dbfe165aaf7c4bddaa10220bd2286db", "generation_seconds_batch": 24.64519719593227, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.105729103088379, "target_z": 1.0937499027280755, "ce": 6.109982967376709, "genuineness": 4.444586753845215, "identity": 0.582192063331604, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0df0c727931b4bcdba28b9929a1d67e74c449c017e3b3f076b2f48f636d75bbb.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "0e71d92a5795b9ed66ba76fac71bb7a01b080bdcd16e1b1bd1752298b425b30f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "90fbbb070ce57273b78b6d486376e10c432667cfa27d9cf8c9c15ec15844b6f5", "generation_seconds_batch": 23.921436959994026, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.4844465255737305, "target_z": 2.0937502485838073, "ce": 5.919358730316162, "genuineness": 4.833701133728027, "identity": 0.4073713421821594, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0e71d92a5795b9ed66ba76fac71bb7a01b080bdcd16e1b1bd1752298b425b30f.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "0e9156bc58bd230ccf89ddee74afba380b10a8bb0506d4b33557c259d8f8b93e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 130, "seconds_generated": 10.4, "mp3_sha256": "2ca79042785ee4f7affb595b61940f559529bc991811bc094f491eee7b8d71c8", "generation_seconds_batch": 24.13006234099157, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.396552324295044, "target_z": 1.3046874864900104, "ce": 5.9431867599487305, "genuineness": 4.369274139404297, "identity": 0.3988404870033264, "hypothesis": "I found the little box. It was beside the old window. Need and look inside a", "wer": 0.1875, "duration_error_s": 0.20000000000000107, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0e9156bc58bd230ccf89ddee74afba380b10a8bb0506d4b33557c259d8f8b93e.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "0f2696a1864655cb68545fa898bf66d4b69036036f412594817e2c1da45fbdc2", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ca3a597aeaaa78011d28caab273b1ab522395fa34840c0479ad324148c99b98e", "generation_seconds_batch": 25.705541954026558, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8256773948669434, "target_z": 0.890625027019979, "ce": 6.38797664642334, "genuineness": 4.21237325668335, "identity": 0.21901482343673706, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0f2696a1864655cb68545fa898bf66d4b69036036f412594817e2c1da45fbdc2.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "0f33b45cd418cef790f070f6d95e3951bf2c805b4c0371bff56d15e4bb81af59", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 130, "seconds_generated": 10.4, "mp3_sha256": "e5c457ab0ccd049b8206c9557886aa28e9392281f568ba08794df18b36dbe8c9", "generation_seconds_batch": 23.86717946291901, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5904345512390137, "target_z": 1.4453125999739225, "ce": 5.816103935241699, "genuineness": 4.419482707977295, "identity": 0.4016026258468628, "hypothesis": "I found the little box. It was beside the old window. We can look inside.", "wer": 0.0625, "duration_error_s": 0.20000000000000107, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0f33b45cd418cef790f070f6d95e3951bf2c805b4c0371bff56d15e4bb81af59.mp3" }, { "id": "heldout-es-4", "language": "es", "sentences": [ "Escribí el número en el sobre.", "Léelo bien... y comprueba las dos últimas cifras.", "Llámame cuando tengas un momento." ], "text": "Escribí el número en el sobre. Léelo bien... y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-4", "key": "0f5563c6ab6270c280b2b25d220efff8091d3d8b34056da2acc64ca729f5020a", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Escribí el número en el sobre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Llámame cuando tengas un momento.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Escribí el número en el sobre.\n(Natural clear speech.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Natural clear speech.) [3.2 seconds duration] Llámame cuando tengas un momento.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "27e1ef8d4e7dd250338f132a2970e76104ef98991e09de0a785a49ebf1f1b6e4", "generation_seconds_batch": 16.72315739293117, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1703567504882812, "target_z": 1.140625113483912, "ce": 5.959072113037109, "genuineness": 4.406930446624756, "identity": 0.46113359928131104, "hypothesis": "Escribí el número en el sobre. Léelo bien y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0f5563c6ab6270c280b2b25d220efff8091d3d8b34056da2acc64ca729f5020a.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "0f814dccfc73a615f84b5a4626abd3b9089b548eb790323ce8da1a10bd5fd0da", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "f778b005c4ca01d9e6e4d8a4ed8bd41d893a187e3d577cd432874d8d8fc0e0fa", "generation_seconds_batch": 25.056232245988213, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4396371841430664, "target_z": 1.3359374540660356, "ce": 6.038498878479004, "genuineness": 4.595211505889893, "identity": 0.734794020652771, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0f814dccfc73a615f84b5a4626abd3b9089b548eb790323ce8da1a10bd5fd0da.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "0fc5e09c8403008dcfc16faea58b0db2327013f8c17d7e66b3329f374fe8f401", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d0df3044ddc3217e29ce796d3bd90dba4dfb1ee73c1a01576dd4f1ca31e4c427", "generation_seconds_batch": 24.162064735079184, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9674272537231445, "target_z": 1.718749945960042, "ce": 5.5539960861206055, "genuineness": 4.632867813110352, "identity": 0.39136290550231934, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay después.", "wer": 0.058823529411764705, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0fc5e09c8403008dcfc16faea58b0db2327013f8c17d7e66b3329f374fe8f401.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "0fec7bf6379a67313315622f1381f304d4a604ca98914964f9eb4e7a1d5311a2", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0153b48f0b63f48c54b9ffe363f24823ec28df523de386f3053714766ffc5ef7", "generation_seconds_batch": 15.236449437914416, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.355191230773926, "target_z": 1.9999998270721342, "ce": 5.935244083404541, "genuineness": 4.6203155517578125, "identity": 0.6494982242584229, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prête.", "wer": 0.05555555555555555, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0fec7bf6379a67313315622f1381f304d4a604ca98914964f9eb4e7a1d5311a2.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "0fef574fe8c46bd2443fb35af8408e96942a17d4b213192c8bb96e02a337abde", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "53b41126d7731028361348d5fb8125afec81140ce43f8b11478fa43e645447cb", "generation_seconds_batch": 23.144209127989598, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.655061721801758, "target_z": 1.4921874648740272, "ce": 5.530168056488037, "genuineness": 4.582659721374512, "identity": 0.5255421996116638, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/0fef574fe8c46bd2443fb35af8408e96942a17d4b213192c8bb96e02a337abde.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "11234dcb49e64771ca43dab2572a2e044862752dfc3cad2dffdf86c665aeadfc", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "43556f66f3234e3ed5d590d336236e5f2b50c56159b05ab8d204c9cef090297a", "generation_seconds_batch": 23.882941075018607, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.13804292678833, "target_z": 1.1171875081059937, "ce": 6.24500846862793, "genuineness": 4.444586753845215, "identity": 0.31248486042022705, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/11234dcb49e64771ca43dab2572a2e044862752dfc3cad2dffdf86c665aeadfc.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "11a46e0f386f41a5c355537ee647a0a4ac06d106f34131e9c97ecd9f07b673c6", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "94d13cff23792b69476cf03ea114e83cba65e280d8dc06d02d8a537cdbd7a73e", "generation_seconds_batch": 22.102732184925117, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.301335334777832, "target_z": 1.960937497298002, "ce": 5.752562999725342, "genuineness": 4.507347106933594, "identity": 0.8196547627449036, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/11a46e0f386f41a5c355537ee647a0a4ac06d106f34131e9c97ecd9f07b673c6.mp3" }, { "id": "heldout-es-0", "language": "es", "sentences": [ "Tu mensaje llegó después de que me fuera.", "Lo leí dos veces... y guardé el teléfono.", "Dime cuándo podemos hablar." ], "text": "Tu mensaje llegó después de que me fuera. Lo leí dos veces... y guardé el teléfono. Dime cuándo podemos hablar.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-0", "key": "12a7b850c68b1e935d69d2b179f3c0df54c4775061935760902fd9f6cbfc7541", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo podemos hablar.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo podemos hablar.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e69163ec699fbd837535ec44a40b2217201a796934bafa3b07c151c9abaf2697", "generation_seconds_batch": 17.511108755948953, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6596921682357788, "target_z": 0.04492190404647747, "ce": 6.316492557525635, "genuineness": 4.369274139404297, "identity": 0.7316728830337524, "hypothesis": "Tu mensaje llegó después de que me fuera. Lo leí dos veces y guardé el teléfono. Dime cuándo podemos hablar.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/12a7b850c68b1e935d69d2b179f3c0df54c4775061935760902fd9f6cbfc7541.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "12af60bd8b885b9d89d02edb176b696b712da84275a70b01d8c5faef4df6c724", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "340f06d8b60b6d012693c8f93fdc552642a600b526f6ace62723681e4fe148e6", "generation_seconds_batch": 24.94929073692765, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8274013996124268, "target_z": 1.6171875081059937, "ce": 5.073464393615723, "genuineness": 4.030367851257324, "identity": 0.44270697236061096, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás liquido.", "wer": 0.06666666666666667, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/12af60bd8b885b9d89d02edb176b696b712da84275a70b01d8c5faef4df6c724.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "13973b99a9fa3b68c42dffd562b26f844883d2e0d795b29c94a4635232224de7", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b5e1e4f59484657e5d89e8eac7dd0b719b05fdad60225b30e6a3172be7a90195", "generation_seconds_batch": 23.50608081498649, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2099933624267578, "target_z": -0.2812500540399581, "ce": 6.1894097328186035, "genuineness": 4.2751336097717285, "identity": 0.6100850105285645, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/13973b99a9fa3b68c42dffd562b26f844883d2e0d795b29c94a4635232224de7.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "1409a341056f83f1e2bb90ed3fb151aeadfb9c69494f5175dcc0ccce9167c14f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8b02e4df1934bf3ae5b9fbc4d1938762072ec1adf1953777ddbf0626a9cdfe59", "generation_seconds_batch": 23.879537864006124, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.827992558479309, "target_z": 0.1669922033742377, "ce": 6.252951145172119, "genuineness": 4.218649387359619, "identity": 0.5604552030563354, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1409a341056f83f1e2bb90ed3fb151aeadfb9c69494f5175dcc0ccce9167c14f.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "14502bb09374b31b15c3eb1461d0ee47d1783e21ae56425ddc30769d5d57c9b1", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9f8fadf70ca55816763f0af431203af10baa9a10fa0730ed07d5a3e8ccc493f9", "generation_seconds_batch": 23.293156858999282, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.2259368896484375, "target_z": 1.9062500972719245, "ce": 5.935244083404541, "genuineness": 4.394378662109375, "identity": 0.36425983905792236, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/14502bb09374b31b15c3eb1461d0ee47d1783e21ae56425ddc30769d5d57c9b1.mp3" }, { "id": "heldout-en-1", "language": "en", "sentences": [ "The keys are on the kitchen table.", "I checked the drawer... then looked beside the window.", "Take them with you when you leave." ], "text": "The keys are on the kitchen table. I checked the drawer... then looked beside the window. Take them with you when you leave.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-1", "key": "14b15a31ead93f3319ba8bf63a009840fdc3f88b7aabe3289d22a8979674803d", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The keys are on the kitchen table.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Take them with you when you leave.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The keys are on the kitchen table.\n(Natural clear speech.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Natural clear speech.) [3.2 seconds duration] Take them with you when you leave.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "51f9518df04199e931679815056d4922c92362c9f0a5bbcfd5f0d31ab684d4cd", "generation_seconds_batch": 23.35967026592698, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3211536407470703, "target_z": 1.249999913536067, "ce": 5.601652145385742, "genuineness": 4.168440818786621, "identity": 0.41265201568603516, "hypothesis": "The keys are on the kitchen table. I checked the drawer, then looked beside the window. Take them with you when you", "wer": 0.043478260869565216, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/14b15a31ead93f3319ba8bf63a009840fdc3f88b7aabe3289d22a8979674803d.mp3" }, { "id": "heldout-de-2", "language": "de", "sentences": [ "Die Straße führt hinter der Brücke nach links.", "Geh weiter... bis zu dem kleinen Holztor.", "Ich warte auf der anderen Seite." ], "text": "Die Straße führt hinter der Brücke nach links. Geh weiter... bis zu dem kleinen Holztor. Ich warte auf der anderen Seite.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-2", "key": "15585f3962d31fd159bb39a755593ab24aa0df87e4725d7772f40811258760f5", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3c166e46b35539583af78d34d6ff1bb6032efdde2738fac423cd85962bb50edc", "generation_seconds_batch": 15.55431785795372, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4217562675476074, "target_z": 0.5976561702910618, "ce": 6.300607204437256, "genuineness": 4.6203155517578125, "identity": 0.6013458967208862, "hypothesis": "Die Straße führt hinter der Brücke nach links. Geh weiter bis zum kleinen Holztor. Ich warte auf der anderen Seite.", "wer": 0.09523809523809523, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/15585f3962d31fd159bb39a755593ab24aa0df87e4725d7772f40811258760f5.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "16136217de63c315f0ec4c272caa71630dd606d6011513e00db7684b858f1e3d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "74c8513ae30f7d7481e2e2780299c41373356a3d4e8706b5cdff31dd2c681fe8", "generation_seconds_batch": 24.43354455300141, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.238645553588867, "target_z": 0.4648437648609885, "ce": 6.284721851348877, "genuineness": 4.381826400756836, "identity": 0.6822869181632996, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/16136217de63c315f0ec4c272caa71630dd606d6011513e00db7684b858f1e3d.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "16596c5c280ad42bfac6cde45adf54bbb17c031072470c73996fefe40f93c856", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.0, "cfg": 2.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "d07b5dfd9307bba1fb041faef1c2fa5f2d492e9253b891c6ce0de56b423debeb", "generation_seconds_batch": 18.013539887033403, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9020447731018066, "target_z": 0.22070316620546804, "ce": 6.0305562019348145, "genuineness": 4.595211505889893, "identity": 0.6362399458885193, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd, zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/16596c5c280ad42bfac6cde45adf54bbb17c031072470c73996fefe40f93c856.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "166e426b57ea2dbdfafd7106d10c94b35f136236404d55beac98d6f362ffbc7a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "abdca293aeb76de492f33af41a191f4eb6941bfdb27a00b9fef3a57ff4520276", "generation_seconds_batch": 24.073619749979116, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.987245798110962, "target_z": 1.0078125351259728, "ce": 6.173524379730225, "genuineness": 4.519899368286133, "identity": 0.5498943328857422, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/166e426b57ea2dbdfafd7106d10c94b35f136236404d55beac98d6f362ffbc7a.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "171d15a20637c1ece107c15b9db347a76daf4b7144f72322146bc72de4fc0324", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "410169f8da6f275c78bd25c7779e4a83d8013de8baf9541f21207db5a38e05ef", "generation_seconds_batch": 25.36465780204162, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6583456993103027, "target_z": 0.0439452939237644, "ce": 6.268836498260498, "genuineness": 4.444586753845215, "identity": 0.3229600787162781, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/171d15a20637c1ece107c15b9db347a76daf4b7144f72322146bc72de4fc0324.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "177d0c3edec9a6fd50982351417d3fbea72ff6f3f2e3161f0e2c5776ed042cb0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7f87af6bd9ed0f7bd1a52c2008dafb5ea77a13041ce6c93a794bbc07cb36135e", "generation_seconds_batch": 23.269659119076096, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.901075839996338, "target_z": 0.9453124270460566, "ce": 6.181467056274414, "genuineness": 4.400654315948486, "identity": 0.6016333103179932, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/177d0c3edec9a6fd50982351417d3fbea72ff6f3f2e3161f0e2c5776ed042cb0.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "17c806696e35c005a053352fe06d2b0f2e56393855c7500a842af0b24bbb1548", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c4ee3c76198a64a99ccfe186101fda272b95e56d36b874b6322289630b64390e", "generation_seconds_batch": 23.830862556933425, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.663731336593628, "target_z": 0.047851561486750786, "ce": 6.3641486167907715, "genuineness": 4.457139015197754, "identity": 0.5605944991111755, "hypothesis": "Ich habe die kleine Schachtel gefunden, sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/17c806696e35c005a053352fe06d2b0f2e56393855c7500a842af0b24bbb1548.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "1845f243fe114ce64a8866484225c73d3f4823ff50f0622faddb5783c9e75fd8", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a0333e256068f1445f9e2a1f2c547f7e0b314ca10a3d17df49026eb957fdbabd", "generation_seconds_batch": 23.028504373971373, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4073235988616943, "target_z": 1.3125000216159832, "ce": 6.117925643920898, "genuineness": 4.507347106933594, "identity": 0.4861859679222107, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1845f243fe114ce64a8866484225c73d3f4823ff50f0622faddb5783c9e75fd8.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "185aee8ac29ffe433a608bd6eee13597a999d28c66d182112dfaf7892cc24f27", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "16eca8be64d46998970319bd3500614c6e55c9aaa9afa84be04aa4e55aeefd0f", "generation_seconds_batch": 22.407040906953625, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.998016834259033, "target_z": 1.0156248973240796, "ce": 5.526196479797363, "genuineness": 4.093128204345703, "identity": 0.5691558122634888, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/185aee8ac29ffe433a608bd6eee13597a999d28c66d182112dfaf7892cc24f27.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "18dc4c425bc9c3108202bbf32664b8365c8d17b4630115cff3e4de056047546d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "207252b05205711feecf8f7b535df31813f13746b85e9814a29fa81de7a91f8c", "generation_seconds_batch": 25.28117792296689, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4180946350097656, "target_z": 1.32031238381409, "ce": 5.95112943649292, "genuineness": 4.3504462242126465, "identity": 0.40853166580200195, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/18dc4c425bc9c3108202bbf32664b8365c8d17b4630115cff3e4de056047546d.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "1972f0969db7926e6adffdf7bd5af5a04a096632d8ec0cf85d52cc9e58ef3f7e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "dede4595cbb82ad0d2fc99d9ce53ead82efadd406f960b14c3db6cb8da37baca", "generation_seconds_batch": 17.77146630291827, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4402284622192383, "target_z": -0.11425776420178745, "ce": 6.284721851348877, "genuineness": 4.218649387359619, "identity": 0.5579942464828491, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1972f0969db7926e6adffdf7bd5af5a04a096632d8ec0cf85d52cc9e58ef3f7e.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "19a1a79a336cf902fc4687994e2d68b9dece6cba35d61c05c69c733954832746", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a59eeeaa182c098c6c94a629bc2898d1dcdcd3ad40115d78cb4c31f28e1338e2", "generation_seconds_batch": 17.92267002305016, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6317954063415527, "target_z": 0.7500000864639329, "ce": 6.284721851348877, "genuineness": 4.6203155517578125, "identity": 0.7493970394134521, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/19a1a79a336cf902fc4687994e2d68b9dece6cba35d61c05c69c733954832746.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "19d45a38be2b0a7d30488f28d4f51aece8d05c2f44e80f9007be581a00fcbde5", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "d65b2d7046211c37eeeae2e45b2b45c21c70f4e859797cdbe947ae9bbb16e8ac", "generation_seconds_batch": 23.348486786009744, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.859714984893799, "target_z": 1.640624940556046, "ce": 5.550024509429932, "genuineness": 4.532451152801514, "identity": 0.5408134460449219, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/19d45a38be2b0a7d30488f28d4f51aece8d05c2f44e80f9007be581a00fcbde5.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "1a3e6e91af7d355bb4c728366b3cb3137274b80f5f42beffe78274cac6241745", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f763be4694c0f12d7bc2ac1024221bf1f9c4a7ac745eafefce0fca6d16f667c5", "generation_seconds_batch": 15.723736896063201, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.182851791381836, "target_z": 1.8749999567680335, "ce": 5.847874641418457, "genuineness": 4.65797233581543, "identity": 0.5157869458198547, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay.", "wer": 0.058823529411764705, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1a3e6e91af7d355bb4c728366b3cb3137274b80f5f42beffe78274cac6241745.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "1a8e3460608eab180748d9b79cc34fe54d132eaaa5c1f994841e47b870ef29aa", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "eb24b492c710320f9409d8fae1d9eb13c8181146b8e3fbdf2a81a9b6024372ee", "generation_seconds_batch": 24.273448113002814, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8414565324783325, "target_z": 0.17675778581777069, "ce": 6.41180419921875, "genuineness": 4.532451152801514, "identity": 0.44849473237991333, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1a8e3460608eab180748d9b79cc34fe54d132eaaa5c1f994841e47b870ef29aa.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "1aef0cea1f6a4a212beea47dafa7250b3d9659932855ced08eef34064336d431", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "3650683c3dea4475dea5c6eb90d9202de12022fdc3ed8a64bf536ad62ca4b67a", "generation_seconds_batch": 22.98898829310201, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.601205587387085, "target_z": 1.4531249621720292, "ce": 5.998785495758057, "genuineness": 4.419482707977295, "identity": 0.6099300384521484, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinfahren.", "wer": 0.0625, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1aef0cea1f6a4a212beea47dafa7250b3d9659932855ced08eef34064336d431.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "1b63a31a00eb945ee83663833e11f2029187d13eb45da1bbf70fb220e4a11e79", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "d59d876ac5e8ce54edbcd5d2981dbdc24e3daaade6350981b90e82f8765a5a33", "generation_seconds_batch": 23.876749251037836, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.601205587387085, "target_z": 1.4531249621720292, "ce": 6.09409761428833, "genuineness": 4.632867813110352, "identity": 0.8313528299331665, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1b63a31a00eb945ee83663833e11f2029187d13eb45da1bbf70fb220e4a11e79.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "1b8a7d2469c5616a10bb1bae95fcb2f5f2636a843de9fdd3bcaf261fa85bfeb2", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5028c78f75683d6d4a050a5e43a453375380e3234148efb76d4ebe52edf0db1e", "generation_seconds_batch": 17.126916994107887, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.224212646484375, "target_z": 1.179687443258044, "ce": 6.054384231567383, "genuineness": 4.362998008728027, "identity": 0.7032490968704224, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1b8a7d2469c5616a10bb1bae95fcb2f5f2636a843de9fdd3bcaf261fa85bfeb2.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "1bb5d1f9f3bb115776d674efc78a422a5e29ed980bd1ee22038aa55696057ae8", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ccf4be19ebec5a931bc4a8d2b9735479ddd57725851e835f8848e9e2e9e04103", "generation_seconds_batch": 24.797911969013512, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6317954063415527, "target_z": 0.7500000864639329, "ce": 6.300607204437256, "genuineness": 4.1998209953308105, "identity": 0.1967446357011795, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1bb5d1f9f3bb115776d674efc78a422a5e29ed980bd1ee22038aa55696057ae8.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "1bf69dddf0c360ae5c933902808bd82ffdccfc9d402837bbf79ba0892ff10ee1", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "af5698e6080d82e0190589d14282d1d80c010fab384e8e1b7ad9cb50e4bb4242", "generation_seconds_batch": 24.37529632996302, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3894426822662354, "target_z": 0.5742187378410094, "ce": 6.070269584655762, "genuineness": 4.532451152801514, "identity": 0.4578952193260193, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1bf69dddf0c360ae5c933902808bd82ffdccfc9d402837bbf79ba0892ff10ee1.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "1c31584ecf93fdc88842d629342eba7221209464c5e7ea155119b82493309e2e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "da78fa6d52483b4f49bbb07665f69082e84be77d9204a61c3c2997ffead32084", "generation_seconds_batch": 24.093101347913034, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5796632766723633, "target_z": 1.4375000648479497, "ce": 5.808161735534668, "genuineness": 4.306513786315918, "identity": 0.49163562059402466, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1c31584ecf93fdc88842d629342eba7221209464c5e7ea155119b82493309e2e.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "1c3f493837a29ce5e5faa87d053019fdce2ff2b3f0fe4e14fe10cfc1aa2ddaff", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "10d1872c68a4efa34c4183babf5bf6692e13eeb05ee0e18b4c4a2485636cd7d0", "generation_seconds_batch": 23.192518897005357, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9118471145629883, "target_z": 0.9531249621720294, "ce": 5.720792293548584, "genuineness": 4.055472373962402, "identity": 0.32070988416671753, "hypothesis": "Ik heb het kleine doosje gevonden, het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1c3f493837a29ce5e5faa87d053019fdce2ff2b3f0fe4e14fe10cfc1aa2ddaff.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "1ca54f40b8e1620e8c63d1e5a41b31e64caf47ae246b28bb8dfd236473be5969", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "156e2569bc7385892d477e57638609d27f2aec5c353eac0321a54c83ecf1225f", "generation_seconds_batch": 23.478353463113308, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3634834289550781, "target_z": -0.169921860814511, "ce": 6.165581703186035, "genuineness": 4.444586753845215, "identity": 0.6579588651657104, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1ca54f40b8e1620e8c63d1e5a41b31e64caf47ae246b28bb8dfd236473be5969.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "1d1442950e83af46a9b867ca862dd19cc1eef502fb6cfd9daedb189ece86ac4e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "909b54b3e8b5b9844d0b9b4fecb093bcc82da212f064480d9d21db236b15b32c", "generation_seconds_batch": 24.176370520028286, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1336259841918945, "target_z": 0.38867180677455293, "ce": 6.0305562019348145, "genuineness": 4.080576419830322, "identity": 0.7516834735870361, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1d1442950e83af46a9b867ca862dd19cc1eef502fb6cfd9daedb189ece86ac4e.mp3" }, { "id": "heldout-es-4", "language": "es", "sentences": [ "Escribí el número en el sobre.", "Léelo bien... y comprueba las dos últimas cifras.", "Llámame cuando tengas un momento." ], "text": "Escribí el número en el sobre. Léelo bien... y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-4", "key": "1d559e6da25f8bbacc3b53d5880369445633ce7ab0125ca7b07093108b8c8de3", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Escribí el número en el sobre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Llámame cuando tengas un momento.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Escribí el número en el sobre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Llámame cuando tengas un momento.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0cfd7bd684f85212a284cea29a4f97d8a7f61cb151a4f763e198d9b92331ec2e", "generation_seconds_batch": 16.237887048977427, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2907776832580566, "target_z": -0.22265621352302828, "ce": 6.316492557525635, "genuineness": 4.457139015197754, "identity": 0.7145856022834778, "hypothesis": "Escribí el número en el sobre, léelo bien, y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1d559e6da25f8bbacc3b53d5880369445633ce7ab0125ca7b07093108b8c8de3.mp3" }, { "id": "heldout-de-1", "language": "de", "sentences": [ "Die Schlüssel liegen auf dem Küchentisch.", "Ich suchte in der Schublade... dann neben dem Fenster.", "Nimm sie mit, wenn du gehst." ], "text": "Die Schlüssel liegen auf dem Küchentisch. Ich suchte in der Schublade... dann neben dem Fenster. Nimm sie mit, wenn du gehst.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-1", "key": "1d829402ab2176d2024ae3e918286a39ec094d9a5299caa0e048ba37912e1579", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f5aad5c86f0ca846bc118d6e9356b2fb17dd62e9ebf9b2756cb28bc6e8bab62d", "generation_seconds_batch": 18.238929986022413, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4271419048309326, "target_z": 0.6015624378540482, "ce": 6.3641486167907715, "genuineness": 4.482243061065674, "identity": 0.5711989402770996, "hypothesis": "Die Schlüssel liegen auf dem Küchentisch. Ich suchte in der Schublade, dann neben dem Fenster. Nimm sie mit, wenn du gehst.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1d829402ab2176d2024ae3e918286a39ec094d9a5299caa0e048ba37912e1579.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "1d9039694e24433f106773d0ee169d5fae392aee4dc320a12ba0edb83abc1774", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "5ced37907b5b9cd8a4c27944fa3983f2120ef855f92791e8534498d3e6c66cad", "generation_seconds_batch": 24.26231643895153, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7843165397644043, "target_z": 1.5859375405299685, "ce": 5.323658466339111, "genuineness": 4.268857479095459, "identity": 0.37476834654808044, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1d9039694e24433f106773d0ee169d5fae392aee4dc320a12ba0edb83abc1774.mp3" }, { "id": "heldout-es-1", "language": "es", "sentences": [ "Las llaves están en la mesa de la cocina.", "Miré en el cajón... y luego junto a la ventana.", "Llévatelas cuando te vayas." ], "text": "Las llaves están en la mesa de la cocina. Miré en el cajón... y luego junto a la ventana. Llévatelas cuando te vayas.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-1", "key": "1e0d81780906f4fec33f8af5e38c1ad9ce3d11fd0d525a9b469edff812387317", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Llévatelas cuando te vayas.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Natural clear speech.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Natural clear speech.) [3.2 seconds duration] Llévatelas cuando te vayas.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1749f460143d89ca87f271a597180fde7f5b3a0bc8eed25c760546f069ca454e", "generation_seconds_batch": 17.253805133979768, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5365781784057617, "target_z": 1.4062499243440587, "ce": 5.982900142669678, "genuineness": 4.570107460021973, "identity": 0.5990243554115295, "hypothesis": "Las llaves están en la mesa de la cocina. Miren el cajón y luego junto a la ventana. Llévatelas cuando te vayas.", "wer": 0.08695652173913043, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1e0d81780906f4fec33f8af5e38c1ad9ce3d11fd0d525a9b469edff812387317.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "1e4d7a5c74c481deaa031248f698935ae7f6a345b4857c0a24cd299ff9136a53", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b486b435f217f1cb12a9809719df3a35f2bd1a589662fa9c7a3df61a16635a6b", "generation_seconds_batch": 24.687736685969867, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7089178562164307, "target_z": 1.5312499675760252, "ce": 6.038498878479004, "genuineness": 4.6203155517578125, "identity": 0.7632695436477661, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1e4d7a5c74c481deaa031248f698935ae7f6a345b4857c0a24cd299ff9136a53.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "1e951c46ee3006bdb9bb6ce70273fa10e9fa37ebb366b55c3e2662745d79b647", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "413bdd4f1513bf7dc1d811735081edcf281acd405e0dee890727d4bb5c9f4219", "generation_seconds_batch": 23.611274659982882, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8791558742523193, "target_z": 0.2041015722947424, "ce": 6.221180438995361, "genuineness": 4.406930446624756, "identity": 0.7936902046203613, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1e951c46ee3006bdb9bb6ce70273fa10e9fa37ebb366b55c3e2662745d79b647.mp3" }, { "id": "heldout-es-4", "language": "es", "sentences": [ "Escribí el número en el sobre.", "Léelo bien... y comprueba las dos últimas cifras.", "Llámame cuando tengas un momento." ], "text": "Escribí el número en el sobre. Léelo bien... y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-4", "key": "1eaf0e0e2c966186cf6ada2de41819a315f8b1b1eeeeef5a84a28dd582998db8", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Escribí el número en el sobre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Llámame cuando tengas un momento.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Escribí el número en el sobre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Llámame cuando tengas un momento.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "939bf00a6f201370474b5acb8456cca4d6b84049f91c2ced3d74023e3be85b98", "generation_seconds_batch": 16.237887048977427, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9487912058830261, "target_z": -0.4707031229735016, "ce": 6.332377910614014, "genuineness": 4.325342178344727, "identity": 0.6331673860549927, "hypothesis": "Escribí el número en el sobre. Léelo bien y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1eaf0e0e2c966186cf6ada2de41819a315f8b1b1eeeeef5a84a28dd582998db8.mp3" }, { "id": "heldout-fr-3", "language": "fr", "sentences": [ "Quelqu’un a déplacé la petite chaise bleue.", "Elle était ici... près de la grande bibliothèque.", "Nous la remettrons après le déjeuner." ], "text": "Quelqu’un a déplacé la petite chaise bleue. Elle était ici... près de la grande bibliothèque. Nous la remettrons après le déjeuner.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-3", "key": "1ec014369eb939264d8c16de01ed11fcb3968e9bfb36e99b447c62ca8e5519dc", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Natural clear speech.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Natural clear speech.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "55a284e9967028b36dd54b34f807de402c65d45639b29966b962e24ee673eb2c", "generation_seconds_batch": 24.642980829929, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5581207275390625, "target_z": 1.421874994596004, "ce": 5.577824115753174, "genuineness": 4.482243061065674, "identity": 0.7277440428733826, "hypothesis": "Quelqu'un a déplacé la petite chaise bleue. Elle était ici, près de la grande bibliothèque. Nous la remettrons après le déjeuner.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1ec014369eb939264d8c16de01ed11fcb3968e9bfb36e99b447c62ca8e5519dc.mp3" }, { "id": "heldout-nl-3", "language": "nl", "sentences": [ "Iemand heeft het kleine blauwe stoeltje verplaatst.", "Het stond hier... naast de hoge boekenkast.", "We zetten het na de lunch terug." ], "text": "Iemand heeft het kleine blauwe stoeltje verplaatst. Het stond hier... naast de hoge boekenkast. We zetten het na de lunch terug.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-3", "key": "1eff5985824ed961e7afc0b3600a97915836e054d6bde7db727f0c4fb3226110", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We zetten het na de lunch terug.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Natural clear speech.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Natural clear speech.) [3.2 seconds duration] We zetten het na de lunch terug.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "88a42bdb928a782757c52c346d0aa6ecc8d8a8ba9adaa272e19b791feb6adc54", "generation_seconds_batch": 23.683933655032888, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.960317611694336, "target_z": 0.9882811973110408, "ce": 6.09409761428833, "genuineness": 4.457139015197754, "identity": 0.6124006509780884, "hypothesis": "Iemand heeft het kleine blauwe stoeltje verplaatst. Het stond hier naast de hoge boekenkast. We zetten het naar de lunch terug.", "wer": 0.047619047619047616, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1eff5985824ed961e7afc0b3600a97915836e054d6bde7db727f0c4fb3226110.mp3" }, { "id": "heldout-fr-2", "language": "fr", "sentences": [ "La route tourne à gauche après le pont.", "Continue... jusqu’à la petite barrière en bois.", "Je t’attendrai de l’autre côté." ], "text": "La route tourne à gauche après le pont. Continue... jusqu’à la petite barrière en bois. Je t’attendrai de l’autre côté.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-2", "key": "1f344bbed70c594cfd5b67de81c6b330a590a956deab3dbc1c7764280b653a54", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Natural clear speech.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Natural clear speech.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "ad2ff50ebf0f6c766a0b0a014f07d605e368bb33e03bbddea974455d1e866943", "generation_seconds_batch": 24.339199988986365, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.139766693115234, "target_z": 1.8437498162641426, "ce": 5.40705680847168, "genuineness": 4.494794845581055, "identity": 0.645383358001709, "hypothesis": "La route tourne à gauche après le pont, continue jusqu'à la petite barrière en bois.", "wer": 0.25, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1f344bbed70c594cfd5b67de81c6b330a590a956deab3dbc1c7764280b653a54.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "1f7c44c7f4c4939cf24a9434a076f4f2298d0994a309042545cd43bb6107550a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "e9762a3ccd70e16f2969a74ca01789956f010e3d8c4804f815b2bb0e0f96af88", "generation_seconds_batch": 23.114793732995167, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.547349452972412, "target_z": 1.4140624594700315, "ce": 6.252951145172119, "genuineness": 4.337893962860107, "identity": 0.35867685079574585, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/1f7c44c7f4c4939cf24a9434a076f4f2298d0994a309042545cd43bb6107550a.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "202d7d69315b135326b91625b38e9cf3bd739c7ed38d049d6fe78f96c6183ecd", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e6adbd3a545ecdfeeeb07c49efaa5a3cc5851dd547aed38cd18c085926ad2226", "generation_seconds_batch": 21.023946276982315, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.924342393875122, "target_z": 1.6874999783840168, "ce": 6.078212261199951, "genuineness": 4.337893962860107, "identity": 0.5936728715896606, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hin.", "wer": 0.0625, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/202d7d69315b135326b91625b38e9cf3bd739c7ed38d049d6fe78f96c6183ecd.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "204cc9b6ebc2e5e3cb2eb0fd33a97e0e463ffeca52d34e18dc9ee89997a67cb2", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "72daec3a6cfc8877a9e17de193bdf2582ccd29df4067e87610de48cf5831d0c1", "generation_seconds_batch": 24.786545202950947, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.053597450256348, "target_z": 1.781250226967824, "ce": 5.804190158843994, "genuineness": 4.758388519287109, "identity": 0.44711291790008545, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/204cc9b6ebc2e5e3cb2eb0fd33a97e0e463ffeca52d34e18dc9ee89997a67cb2.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "2066d43790e6a1b9746dc92a8eccb7d26d3c73ef4b9fb5aaf774109440ef202c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "fb75a4c62c024357a3ce03d5110d46f97942b1a615ce212616cb2a745a7cc0c2", "generation_seconds_batch": 18.202271878020838, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.364238739013672, "target_z": 1.281250054039958, "ce": 5.80021858215332, "genuineness": 4.281409740447998, "identity": 0.5896257162094116, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2066d43790e6a1b9746dc92a8eccb7d26d3c73ef4b9fb5aaf774109440ef202c.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "20a295a150be3afd083aec05fe4d1d381cd2b04d27b9806160a45028e7ffd263", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3170f2d740db333a781e9b3a3a765f77d68052c92713a98365d3cc09c2a9b748", "generation_seconds_batch": 17.492625884944573, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.924342393875122, "target_z": 1.6874999783840168, "ce": 6.078212261199951, "genuineness": 4.457139015197754, "identity": 0.5791148543357849, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/20a295a150be3afd083aec05fe4d1d381cd2b04d27b9806160a45028e7ffd263.mp3" }, { "id": "heldout-de-0", "language": "de", "sentences": [ "Deine Nachricht kam, als ich schon weg war.", "Ich las sie zweimal... dann legte ich das Handy weg.", "Sag mir, wann wir reden können." ], "text": "Deine Nachricht kam, als ich schon weg war. Ich las sie zweimal... dann legte ich das Handy weg. Sag mir, wann wir reden können.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-0", "key": "20d7c6e25598bec8956d5a59c28510e9627ca8d43dc0f4bc0da6753af5638798", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wann wir reden können.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Natural clear speech.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wann wir reden können.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9c213d746a9363f71ec76c326aedc2754ede7871abdc880b3a6ca908d6a8dbc1", "generation_seconds_batch": 15.556458711973391, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6119768619537354, "target_z": 1.460937497298002, "ce": 5.6413655281066895, "genuineness": 4.532451152801514, "identity": 0.6060724258422852, "hypothesis": "Deine Nachricht kam, als ich schon weg war. Ich las sie zweimal, dann legte ich das Handy weg. Sag mir, wann wir reden können.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/20d7c6e25598bec8956d5a59c28510e9627ca8d43dc0f4bc0da6753af5638798.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "2129f7b7d22dca99d58000509809acd930833298ce17b164508efc65152df37a", "bucket": "vn_S_ASMR_reg", "phase": "wording", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "14020b555e1f307b7415adf1a7768e2f8a2e90254ac8b05e0d443c7e5c8c1d1f", "generation_seconds_batch": 17.368578661000356, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8185677528381348, "target_z": 0.160156278370978, "ce": 6.41180419921875, "genuineness": 4.469690799713135, "identity": 0.6604843139648438, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2129f7b7d22dca99d58000509809acd930833298ce17b164508efc65152df37a.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "21ee9f1b07a165fd59b2640b394e30f1d331541384e042e8a7c39b014838f04f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "74c21cd8838f60cdd064f66e8f304cbde3a00bbd031cc9aa3c84c8a51b014d7e", "generation_seconds_batch": 19.000402122968808, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3211536407470703, "target_z": 1.249999913536067, "ce": 6.141753673553467, "genuineness": 4.7458367347717285, "identity": 0.6937966346740723, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/21ee9f1b07a165fd59b2640b394e30f1d331541384e042e8a7c39b014838f04f.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "223ed1bcfd73cc8349a43eaaa7bab8de0fd9213155e6df59f4d0ad165e74ada6", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a187123d835e4f2a2b5957fdd08a044718a61c390da1694e3ff7294cccaa6010", "generation_seconds_batch": 22.516669878037646, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.364238739013672, "target_z": 1.281250054039958, "ce": 6.117925643920898, "genuineness": 4.293961524963379, "identity": 0.6398595571517944, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/223ed1bcfd73cc8349a43eaaa7bab8de0fd9213155e6df59f4d0ad165e74ada6.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "2266b02db2a0e17decdc372d6578621dfb4fe25d5a6b3627ec2b63a23e815131", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f3f25a9a699cacd2c04eafe39ab27243ff6727171cb8212335c64f95f05523b6", "generation_seconds_batch": 24.51385081699118, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8041348457336426, "target_z": 0.8749999567680335, "ce": 5.974957466125488, "genuineness": 4.369274139404297, "identity": 0.6367872953414917, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2266b02db2a0e17decdc372d6578621dfb4fe25d5a6b3627ec2b63a23e815131.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "22945c572bb6d63026bfdfad3269622a95efde63fcb35426b7418b986ad47554", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "fc0600887462096ef796ff6e246475a91e2b9f0391a8cc597bdb928d9fb1da45", "generation_seconds_batch": 24.60167984396685, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.242307186126709, "target_z": -0.2578124486620398, "ce": 6.173524379730225, "genuineness": 4.557555198669434, "identity": 0.6320717930793762, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/22945c572bb6d63026bfdfad3269622a95efde63fcb35426b7418b986ad47554.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "229c31d63c7bb971ca859e522d0cdb9103137e013a42be1cde70907a9850083c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4b99b32e059c14d8237e2ae5c4f85b1a991334726911a7d4a2709d8f9da577e6", "generation_seconds_batch": 23.25922657107003, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.290564060211182, "target_z": 1.9531249621720292, "ce": 5.677107334136963, "genuineness": 4.507347106933594, "identity": 0.7937643527984619, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/229c31d63c7bb971ca859e522d0cdb9103137e013a42be1cde70907a9850083c.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "22b1d8c80a806e4093e6dcdcea087697121dc07a29784b6c6af2757482f3e151", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6137526444ed84902c579ddf7f32d862db98918231b09042a07f5bfda198de27", "generation_seconds_batch": 23.32810746098403, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.051873207092285, "target_z": 1.0546875729539433, "ce": 6.054384231567383, "genuineness": 4.362998008728027, "identity": 0.6406579613685608, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/22b1d8c80a806e4093e6dcdcea087697121dc07a29784b6c6af2757482f3e151.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "22e5db18c295fe65574d9303b8df4bbb77653f700298a5f47db00fc151b2729d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "aaa49f07566d60fffbf99aabc8aa97d4f6bb1af38954136033b89c10b4d4f68a", "generation_seconds_batch": 17.579466392984614, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5008165836334229, "target_z": -0.07031247027802305, "ce": 5.998785495758057, "genuineness": 3.653805732727051, "identity": 0.3947807550430298, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás lista.", "wer": 0.06666666666666667, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/22e5db18c295fe65574d9303b8df4bbb77653f700298a5f47db00fc151b2729d.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "2369f9a3033f996b9ca3a1a161fe4ffe8064f7663ed6a918984f2d7b44f3d183", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4b26b5d6354926a9482c422d66c61874b516f08b5daaf89f9c84862fbd6ba364", "generation_seconds_batch": 21.150289794895798, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.319429874420166, "target_z": 0.5234376053779183, "ce": 6.268836498260498, "genuineness": 4.457139015197754, "identity": 0.3586413860321045, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2369f9a3033f996b9ca3a1a161fe4ffe8064f7663ed6a918984f2d7b44f3d183.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "237d096bbd057219a945b316d3157a0a13d49fcd60f6cd9dbbcbc4b55effb699", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3e881b9e2f2fffd68a959b07faeff0eeea94ef9af54c0b3d483d5a1aea3de879", "generation_seconds_batch": 24.952836701995693, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.093234062194824, "target_z": 0.3593750594439539, "ce": 6.340320587158203, "genuineness": 4.7332844734191895, "identity": 0.4054220914840698, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/237d096bbd057219a945b316d3157a0a13d49fcd60f6cd9dbbcbc4b55effb699.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "241e27f97e4f7dbd6126e74a8807bb3c94379d6c8a5dcce6e741cc24c934e60a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f72aae93db2b8c3279081038a7b3dbb1fa9f8f5b91dedc87972d9e6274b761bc", "generation_seconds_batch": 26.599157833959907, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4180946350097656, "target_z": 1.32031238381409, "ce": 5.879645347595215, "genuineness": 4.268857479095459, "identity": 0.6214718222618103, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschalten.", "wer": 0.0625, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/241e27f97e4f7dbd6126e74a8807bb3c94379d6c8a5dcce6e741cc24c934e60a.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "24782cf36dbe50a581225a29b954fe233107b78c3c6f1e76f08e02f6b1cbfed5", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8672ea016cdb4c158fe7fc7c2d3c47777bf435485a9f841228869d476f33b8b4", "generation_seconds_batch": 17.492625884944573, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4180946350097656, "target_z": 1.32031238381409, "ce": 6.157639026641846, "genuineness": 4.055472373962402, "identity": 0.232596755027771, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/24782cf36dbe50a581225a29b954fe233107b78c3c6f1e76f08e02f6b1cbfed5.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "25416ba90d966dbd450fd7c046829da7604986be5b8c453e896c6a6b9d1e4898", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "680f19b7883568815c11284e3adcafd80a1bd5bdb98af972f822561bcab5032d", "generation_seconds_batch": 23.98545291798655, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0411019325256348, "target_z": 1.0468750378279708, "ce": 6.165581703186035, "genuineness": 4.532451152801514, "identity": 0.5726085901260376, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/25416ba90d966dbd450fd7c046829da7604986be5b8c453e896c6a6b9d1e4898.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "25941d8af3a180c2909bc92edce291f7880c3c31331e0b052725de944f05d698", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "52be6094ba5999f3b1cb75a288470774be29b3e5d77d5a3c472176ce953011ce", "generation_seconds_batch": 24.43250141700264, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.07513952255249, "target_z": 1.7968749513640376, "ce": 5.808161735534668, "genuineness": 4.469690799713135, "identity": 0.5396645665168762, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/25941d8af3a180c2909bc92edce291f7880c3c31331e0b052725de944f05d698.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "259cc906db6c4928d97ed63cc2456d62c0ba4415907819e6cdc6bed501ce6e65", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5b410e3fe25a70a8546f666480dd339d6e2402b10ae864fb619452596e024e4e", "generation_seconds_batch": 24.232628850033507, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.592158317565918, "target_z": 2.171874908132071, "ce": 5.685050010681152, "genuineness": 4.645420074462891, "identity": 0.7956108450889587, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/259cc906db6c4928d97ed63cc2456d62c0ba4415907819e6cdc6bed501ce6e65.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "259d52763f2bf784aa78c0e95b3ddafdb1d978b2667ff4543d80c3a84d4326b0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a23c9efce120e089ce5f2c4a99cf820bae7a9d0409a300b0be0546a45bbe87d6", "generation_seconds_batch": 24.51385081699118, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1703567504882812, "target_z": 1.140625113483912, "ce": 5.752562999725342, "genuineness": 4.432034492492676, "identity": 0.5863234400749207, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/259d52763f2bf784aa78c0e95b3ddafdb1d978b2667ff4543d80c3a84d4326b0.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "26675ab113cb496ccf5101081ad587b128c899b4b93e7ac0ff71df332835a0c1", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "898e7bb6b97ebccda5c8c45bcfbc17f118fe84b622b3e8c30c99975a77937106", "generation_seconds_batch": 21.023946276982315, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.428865909576416, "target_z": 1.3281249189400628, "ce": 6.054384231567383, "genuineness": 4.532451152801514, "identity": 0.5831912755966187, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinspringen.", "wer": 0.0625, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/26675ab113cb496ccf5101081ad587b128c899b4b93e7ac0ff71df332835a0c1.mp3" }, { "id": "heldout-en-1", "language": "en", "sentences": [ "The keys are on the kitchen table.", "I checked the drawer... then looked beside the window.", "Take them with you when you leave." ], "text": "The keys are on the kitchen table. I checked the drawer... then looked beside the window. Take them with you when you leave.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-1", "key": "2687e2b84a53ce77fdcc020240647053547be7c67b42d74da829f11de476bc55", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The keys are on the kitchen table.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Take them with you when you leave.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The keys are on the kitchen table.\n(Natural clear speech.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Natural clear speech.) [3.2 seconds duration] Take them with you when you leave.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6d19aca9e018ccbf2e6bf4864f974d6a2cd4428cfc46050ab47f3d2f13888a1d", "generation_seconds_batch": 17.757892654975876, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.51503586769104, "target_z": 1.3906250270199791, "ce": 5.788304805755615, "genuineness": 4.218649387359619, "identity": 0.38895365595817566, "hypothesis": "The keys are on the kitchen table. I checked the drawer, then looked beside the window. Take them with you when you leave.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2687e2b84a53ce77fdcc020240647053547be7c67b42d74da829f11de476bc55.mp3" }, { "id": "heldout-en-1", "language": "en", "sentences": [ "The keys are on the kitchen table.", "I checked the drawer... then looked beside the window.", "Take them with you when you leave." ], "text": "The keys are on the kitchen table. I checked the drawer... then looked beside the window. Take them with you when you leave.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-1", "key": "26d12995a5b024b42cbe037fdee48ffb67b79df656c09f059592185aa881aca9", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The keys are on the kitchen table.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Take them with you when you leave.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The keys are on the kitchen table.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Take them with you when you leave.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6590b28212a519b0a0eb3dcea02db133eb3fee718fd10d4620e45480ce3a723e", "generation_seconds_batch": 15.52945992606692, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3702154159545898, "target_z": -0.1650390695927445, "ce": 6.435632705688477, "genuineness": 4.293961524963379, "identity": 0.15336915850639343, "hypothesis": "The keys are on the kitchen table. I checked the drawer, then looked beside the window. Take them with you when you leave.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/26d12995a5b024b42cbe037fdee48ffb67b79df656c09f059592185aa881aca9.mp3" }, { "id": "heldout-fr-4", "language": "fr", "sentences": [ "J’ai écrit le numéro sur l’enveloppe.", "Lis-le bien... et vérifie les deux derniers chiffres.", "Appelle-moi quand tu auras un moment." ], "text": "J’ai écrit le numéro sur l’enveloppe. Lis-le bien... et vérifie les deux derniers chiffres. Appelle-moi quand tu auras un moment.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-4", "key": "2703f91f7f7baed9253283ede29fefb8886ce1ad5ddc63a4d9dcf5164b5c07a5", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Natural clear speech.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Natural clear speech.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ec676b534161e1c9043a5f8dd5a099b8516603b818dfb4814320f9fdd5e4df89", "generation_seconds_batch": 15.1527314389823, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3857812881469727, "target_z": 1.2968751242919037, "ce": 5.474569320678711, "genuineness": 4.344170093536377, "identity": 0.7917202711105347, "hypothesis": "J'ai écrit le numéro sur l'enveloppe, le bien, et vérifie les deux derniers chiffres. Appelle-moi quand tu auras un moment.", "wer": 0.045454545454545456, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2703f91f7f7baed9253283ede29fefb8886ce1ad5ddc63a4d9dcf5164b5c07a5.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "27111fb88224e64279c2c35148133fd4d276021583aa2a6409c55b4cf13b4962", "bucket": "vn_S_ASMR_reg", "phase": "wording", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ddf6c558cd3159c6c0c97b5a60aadb138ceec4777fa7a7ea24f43945e249f422", "generation_seconds_batch": 17.368578661000356, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3850258588790894, "target_z": -0.15429687702649844, "ce": 6.292664527893066, "genuineness": 4.444586753845215, "identity": 0.6637420654296875, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/27111fb88224e64279c2c35148133fd4d276021583aa2a6409c55b4cf13b4962.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "2786a5480a51adce3ba73336efabeacf92fe72538426722d0f040682056d0f57", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5a530e36cfd5e0c6194dc6e5e8ca16b07a7e620e9872af41d76d3f822f7dd257", "generation_seconds_batch": 23.403526039095595, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7445155382156372, "target_z": 0.10644531553974765, "ce": 6.427689552307129, "genuineness": 4.369274139404297, "identity": 0.25470417737960815, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2786a5480a51adce3ba73336efabeacf92fe72538426722d0f040682056d0f57.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "279081d085da04a9a890056743e08956184502b172d80fcc83f6a293ae6226b8", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "187a63a990c28584ab54f1befe2f4285eb3a761e6f9dac782992de7e15019d69", "generation_seconds_batch": 23.10274878994096, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.568892002105713, "target_z": 1.4296875297219769, "ce": 6.006728172302246, "genuineness": 4.507347106933594, "identity": 0.7602903842926025, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/279081d085da04a9a890056743e08956184502b172d80fcc83f6a293ae6226b8.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "279b735ef97668dfc26210b45e6fb7b05f79501bc49c1dbd3fdbebb3b70453ad", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5c28fc5ab75dbe52e6da66abc9a320fff4f6ee42923f59f6e3f4be8ee7cf15b8", "generation_seconds_batch": 16.808574029942974, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4934933185577393, "target_z": 1.3749999567680335, "ce": 5.959072113037109, "genuineness": 4.645420074462891, "identity": 0.6065546274185181, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/279b735ef97668dfc26210b45e6fb7b05f79501bc49c1dbd3fdbebb3b70453ad.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "27c47a37f5bdfaed331564a46eec4c95a3f7d300c81dff27305447ce9988a200", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "262a69f936686f0cc05cdf8c46fe89eb109aa69dad9feed0cf05d12b3bbfee86", "generation_seconds_batch": 16.690557966940105, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8579909801483154, "target_z": 0.9140624594700314, "ce": 6.38797664642334, "genuineness": 4.557555198669434, "identity": 0.7627643942832947, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/27c47a37f5bdfaed331564a46eec4c95a3f7d300c81dff27305447ce9988a200.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "288b92d242be56da1b63abd8804a9b8d9d281967acac132b9f678f2e262067d6", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4d083b20ab58e574d7929e32dfa86220ed92cde5534417a18cf740b3a208297c", "generation_seconds_batch": 23.921436959994026, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.118224143981934, "target_z": 1.828124746012197, "ce": 5.804190158843994, "genuineness": 4.632867813110352, "identity": 0.49075591564178467, "hypothesis": "He encontrado la cajita, estaba junto a la ventana vieja. Podemos mirar juntos lo que hay detrás.", "wer": 0.058823529411764705, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/288b92d242be56da1b63abd8804a9b8d9d281967acac132b9f678f2e262067d6.mp3" }, { "id": "heldout-es-0", "language": "es", "sentences": [ "Tu mensaje llegó después de que me fuera.", "Lo leí dos veces... y guardé el teléfono.", "Dime cuándo podemos hablar." ], "text": "Tu mensaje llegó después de que me fuera. Lo leí dos veces... y guardé el teléfono. Dime cuándo podemos hablar.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-0", "key": "28e637cd5feeef931d17764dfbc2eb070a3ebb18b0e49b3fec82438fd516dace", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo podemos hablar.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Natural clear speech.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo podemos hablar.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d59bafa4e1881d5355eed1c737440f0558e9d580e55a1331ddb4fb19f6e9033e", "generation_seconds_batch": 17.429731165990233, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8166301250457764, "target_z": 1.6093749729800209, "ce": 5.371314525604248, "genuineness": 4.469690799713135, "identity": 0.5034294128417969, "hypothesis": "Tu mensaje llegó después de que me fuera, lo leí dos veces y guardé el teléfono. Dime cuándo podemos hablar.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/28e637cd5feeef931d17764dfbc2eb070a3ebb18b0e49b3fec82438fd516dace.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "297c343de50a18b7d22ff387b7cba16799175a629af61c8eecfb466510e057f3", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a61735aef1c1085d36a1bd35a0a2f5d2c95e50959e1a7e1a7750d034ff48c687", "generation_seconds_batch": 23.06798124301713, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.428865909576416, "target_z": 1.3281249189400628, "ce": 6.24500846862793, "genuineness": 4.607763767242432, "identity": 0.7026991844177246, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/297c343de50a18b7d22ff387b7cba16799175a629af61c8eecfb466510e057f3.mp3" }, { "id": "heldout-es-1", "language": "es", "sentences": [ "Las llaves están en la mesa de la cocina.", "Miré en el cajón... y luego junto a la ventana.", "Llévatelas cuando te vayas." ], "text": "Las llaves están en la mesa de la cocina. Miré en el cajón... y luego junto a la ventana. Llévatelas cuando te vayas.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-1", "key": "2a1b2756acef054abb7841a2079e667281b0b8399cb340439aacae53555293d5", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Llévatelas cuando te vayas.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Natural clear speech.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Natural clear speech.) [3.2 seconds duration] Llévatelas cuando te vayas.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "bc879ab282ee13778e6f8ec229a30bbae9e754a4a0eb4744d0f5b291459040ba", "generation_seconds_batch": 24.245687233982608, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.461179733276367, "target_z": 1.3515625243179812, "ce": 5.669164657592773, "genuineness": 4.557555198669434, "identity": 0.45393919944763184, "hypothesis": "Las llaves están en la mesa de la cocina. Mire en el cajón y luego junto a la ventana, llévatelas cuando te vayas.", "wer": 0.043478260869565216, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2a1b2756acef054abb7841a2079e667281b0b8399cb340439aacae53555293d5.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "2af70250e2bf682887e96560d9bc2e5b264e248b0294d5027708fe39ac8367ca", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5abcf9763366c5f9902dc4704b2f2a98c2f2949e3226907875ff323310d5b238", "generation_seconds_batch": 25.206191510078497, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.032054901123047, "target_z": 1.7656251567158785, "ce": 5.577824115753174, "genuineness": 4.532451152801514, "identity": 0.5639968514442444, "hypothesis": "He encontrado la cajita, estaba junto a la ventana vieja, podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2af70250e2bf682887e96560d9bc2e5b264e248b0294d5027708fe39ac8367ca.mp3" }, { "id": "heldout-de-4", "language": "de", "sentences": [ "Ich schrieb die Nummer auf den Umschlag.", "Lies sie genau... und prüfe die letzten beiden Ziffern.", "Ruf an, wenn du Zeit hast." ], "text": "Ich schrieb die Nummer auf den Umschlag. Lies sie genau... und prüfe die letzten beiden Ziffern. Ruf an, wenn du Zeit hast.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-4", "key": "2b2fc0c5f06e1b7e4f51674db07a41b621852dc0bad3c04ab4c999e3f04e64d6", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Natural clear speech.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Natural clear speech.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "df55865a40bf73bf0b5395a1c5d3c39856a48e8fb956daca5f69b4a8fd505eab", "generation_seconds_batch": 16.163580458029173, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.601205587387085, "target_z": 1.4531249621720292, "ce": 5.911416053771973, "genuineness": 4.457139015197754, "identity": 0.6442216634750366, "hypothesis": "Ich schrieb die Nummer auf den Umschlag, ließ sie genau und prüfe die letzten beiden Ziffern. Ruf an, wenn du Zeit hast.", "wer": 0.045454545454545456, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2b2fc0c5f06e1b7e4f51674db07a41b621852dc0bad3c04ab4c999e3f04e64d6.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "2b4b6b9e7f441f5ae43d7239b9d70c4566d5bd6e0f8d8a89e9afc50704fcf62f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "0c7a1cf432675fbf186d991af56e9da69db493d008fe4da2fe468c68cfbf95b8", "generation_seconds_batch": 24.02223238791339, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.978198528289795, "target_z": 1.7265624810860147, "ce": 5.442798614501953, "genuineness": 4.400654315948486, "identity": 0.4118349850177765, "hypothesis": "He encontrado la cajita, estaba junto a la ventana vieja, podemos mirar juntos.", "wer": 0.23529411764705882, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2b4b6b9e7f441f5ae43d7239b9d70c4566d5bd6e0f8d8a89e9afc50704fcf62f.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "2b63b3f9d34bf5aaa78e11aca92af7f2523adebfb232d6e61c07acb6b04ad87a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "70bc891f83a268bcb1b6709f20bc16e7952b111a2b1c5085d12db4e2b439d70b", "generation_seconds_batch": 22.975548913003877, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.529468536376953, "target_z": 0.6757811756950576, "ce": 6.475345611572266, "genuineness": 4.557555198669434, "identity": 0.43454253673553467, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2b63b3f9d34bf5aaa78e11aca92af7f2523adebfb232d6e61c07acb6b04ad87a.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "2b709f6343f926d5c14f3afc42aa48d2038a9c32b90acb348b4a9144e27703a9", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "33dc6f35fa3cb15816a2c319839c8bfcbdc7ef6edac44259593fbb78d90e5bd4", "generation_seconds_batch": 23.951555098989047, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4540700912475586, "target_z": 0.62109377566898, "ce": 6.332377910614014, "genuineness": 4.21237325668335, "identity": 0.25748005509376526, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2b709f6343f926d5c14f3afc42aa48d2038a9c32b90acb348b4a9144e27703a9.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "2b7194de8c04b71be0994111db517abb8f5824b1e8093a171c4dc143861b80a6", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9de9a9423bee933e33333af95b64d9a61d9b605509dae682d8b6a83cdb7b1701", "generation_seconds_batch": 23.830862556933425, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.459455728530884, "target_z": 0.6250000432319665, "ce": 6.475345611572266, "genuineness": 4.507347106933594, "identity": 0.5228314399719238, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2b7194de8c04b71be0994111db517abb8f5824b1e8093a171c4dc143861b80a6.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "2b7294eaa702ef52fce813d86c8c3ec2d685b10d1234d4c37070debf7174e7dd", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "07190569259a4ddb9af744769b655affba4e116ca5c9584e83d84ef9ed2b17dc", "generation_seconds_batch": 26.638734198058955, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1811277866363525, "target_z": 1.1484374756820188, "ce": 6.252951145172119, "genuineness": 4.369274139404297, "identity": 0.4401928186416626, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2b7294eaa702ef52fce813d86c8c3ec2d685b10d1234d4c37070debf7174e7dd.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "2be7f51f7319b3e549e5a2ed41a1620911d4ddc21a049cc38a71ce9a1f34e3e4", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "96f99ff027d2d4d6efe0449d462e2654056716c5581ce7813031e4a39eee660a", "generation_seconds_batch": 17.277718389057554, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1703567504882812, "target_z": 1.140625113483912, "ce": 5.967014789581299, "genuineness": 4.256305694580078, "identity": 0.6237580180168152, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás lista.", "wer": 0.06666666666666667, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2be7f51f7319b3e549e5a2ed41a1620911d4ddc21a049cc38a71ce9a1f34e3e4.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "2c040f520f0693bc3006524f852930ce6148dd1712454186c3728943bb26b05a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c507329e2bdae8e418089aee841c79bd757fd9b73ae40749ff314f7e238d12be", "generation_seconds_batch": 16.92559107101988, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.696422576904297, "target_z": 0.7968749513640377, "ce": 6.260893821716309, "genuineness": 4.582659721374512, "identity": 0.809039831161499, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2c040f520f0693bc3006524f852930ce6148dd1712454186c3728943bb26b05a.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "2c6f3fe7c0424541164c862b60f4aec5c7bfa62205dc45c5be7108f32eb8969a", "bucket": "vn_S_ASMR_reg", "phase": "wording", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "103eaa94a9d764192f0f1ba34904a864e596a7f638712b6817e54ac0bababc45", "generation_seconds_batch": 18.214279919979163, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8633766174316406, "target_z": 0.9179687270330178, "ce": 6.133810997009277, "genuineness": 4.632867813110352, "identity": 0.7390139102935791, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2c6f3fe7c0424541164c862b60f4aec5c7bfa62205dc45c5be7108f32eb8969a.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "2c834e356154169513fd3ebfa054ccd076d532076cce39a6100d263e911ea398", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0b64f801593847656fd90711d3b28e6fb1a54762b4e56308e57ad64d8142192f", "generation_seconds_batch": 16.971309890039265, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4073235988616943, "target_z": 1.3125000216159832, "ce": 5.998785495758057, "genuineness": 4.494794845581055, "identity": 0.6383112668991089, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2c834e356154169513fd3ebfa054ccd076d532076cce39a6100d263e911ea398.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "2d05831a031725d4c0540ae3a0437a82a61f616983208d1605f55dc35386dec3", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5495a8b570c38998f61c8f0c640ac04d24bfd4c400ee51227ecde85344bd65d3", "generation_seconds_batch": 27.643583735916764, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.570615768432617, "target_z": 2.1562498378801256, "ce": 5.673136234283447, "genuineness": 4.482243061065674, "identity": 0.7113543152809143, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle conduit.", "wer": 0.05263157894736842, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2d05831a031725d4c0540ae3a0437a82a61f616983208d1605f55dc35386dec3.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "2d73cfa9c95788111e48390c0092fa3a797468210bbf37affa7d1ede0c20e782", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9e8086805afa0ce6a65f88a3f37c8f78c3bc622d4bdc03d6a07f21af67c523de", "generation_seconds_batch": 26.638734198058955, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1488142013549805, "target_z": 1.1250000432319665, "ce": 6.348263263702393, "genuineness": 4.469690799713135, "identity": 0.3377380073070526, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2d73cfa9c95788111e48390c0092fa3a797468210bbf37affa7d1ede0c20e782.mp3" }, { "id": "heldout-nl-1", "language": "nl", "sentences": [ "De sleutels liggen op de keukentafel.", "Ik keek in de la... en daarna naast het raam.", "Neem ze mee als je weggaat." ], "text": "De sleutels liggen op de keukentafel. Ik keek in de la... en daarna naast het raam. Neem ze mee als je weggaat.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-1", "key": "2e5adeffb1e1736e7b01050e6191e4e3ab966f91db10d9db8760efc7baaee3fc", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Neem ze mee als je weggaat.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Natural clear speech.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Natural clear speech.) [3.2 seconds duration] Neem ze mee als je weggaat.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3f86f55ffeec760f04094e12a52a0d1392cfdcc0293258ad601fdc7fba739d60", "generation_seconds_batch": 25.929826888954267, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.428865909576416, "target_z": 1.3281249189400628, "ce": 5.696964263916016, "genuineness": 4.1998209953308105, "identity": 0.11454333364963531, "hypothesis": "De sleutels liggen op de keukentafel. Ik keek in de la... en daarna, naast het raam, neem ze mee als je weggaat.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2e5adeffb1e1736e7b01050e6191e4e3ab966f91db10d9db8760efc7baaee3fc.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "2e81177f72845d28527cdabf2d1c86882a7fe70e30b61edb02b0843b14de9577", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "a3744e03e28705dcc5105373a12a288388d9aa468e1d90a6cd460e55dfe97367", "generation_seconds_batch": 16.066815444966778, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.441361427307129, "target_z": 2.062500108079916, "ce": 5.577824115753174, "genuineness": 4.494794845581055, "identity": 0.6974285840988159, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce quelque chose.", "wer": 0.10526315789473684, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2e81177f72845d28527cdabf2d1c86882a7fe70e30b61edb02b0843b14de9577.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "2e902d233001e5aa60be6da910e5c5ac544d509f3968b8c0abf695394fd58264", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 130, "seconds_generated": 10.4, "mp3_sha256": "fac388e7f5cdbb71e782edfbae2a7a4d281762466d201276d56d6fbbd4cb3d56", "generation_seconds_batch": 21.895973603939638, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5365781784057617, "target_z": 1.4062499243440587, "ce": 5.355429172515869, "genuineness": 4.18099308013916, "identity": 0.4496667981147766, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.20000000000000107, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2e902d233001e5aa60be6da910e5c5ac544d509f3968b8c0abf695394fd58264.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "2ed6ed90aa48234c3fadc00529ae4a2520956094e4ac33e3f7cb093bcd7b912f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "fd1780c3da81bf0ccedfd856e46515594c1537f72732e06113cc8d08a7666081", "generation_seconds_batch": 24.908641059999354, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3857812881469727, "target_z": 1.2968751242919037, "ce": 6.23706579208374, "genuineness": 4.293961524963379, "identity": 0.3143730163574219, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2ed6ed90aa48234c3fadc00529ae4a2520956094e4ac33e3f7cb093bcd7b912f.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "2ed99003d8adedbb0025f0bfce43e338cfdae0a2642be73549bc6cd2c54afe0c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e9dec2a515a690f9fc117f324e0f9ff9c3fbc2bd7a8a649ca3ad83e368c911cc", "generation_seconds_batch": 17.582223370089196, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.07513952255249, "target_z": 1.7968749513640376, "ce": 5.780362129211426, "genuineness": 4.532451152801514, "identity": 0.6237871646881104, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2ed99003d8adedbb0025f0bfce43e338cfdae0a2642be73549bc6cd2c54afe0c.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "2ef5ccee4a713ee4f9ea457a8d66f1abf86b7c949676cf17412411823cc2c14e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f517f64e581b14a5d8e181fc9ab1ceda5119a07ea18063a934e1ddd971335aa7", "generation_seconds_batch": 22.783314252039418, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.019559383392334, "target_z": 1.0312499675760252, "ce": 6.1894097328186035, "genuineness": 4.519899368286133, "identity": 0.7949419021606445, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2ef5ccee4a713ee4f9ea457a8d66f1abf86b7c949676cf17412411823cc2c14e.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "2f82a7806643896a18c501d6e2e59c1dafecbc7c6b032a170c8592cd2d79d05a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4a908d2d8017d9b5034a74f069cc8ba6ef19776e9b04137b2af545cbc978acd2", "generation_seconds_batch": 23.049930661916733, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2349839210510254, "target_z": 1.1874999783840168, "ce": 6.181467056274414, "genuineness": 4.507347106933594, "identity": 0.6663053035736084, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/2f82a7806643896a18c501d6e2e59c1dafecbc7c6b032a170c8592cd2d79d05a.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "30993bc7fe29a1ee18d1a12093686a601f260f825ac736fd2d009ea3ba8380cc", "bucket": "vn_S_ASMR_reg", "phase": "wording", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "5b765d1ed6d54239ac0a37b032fc3858fd9cb18993bbf8697de434878d2a418b", "generation_seconds_batch": 18.214279919979163, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3625144958496094, "target_z": 0.5546874000260775, "ce": 6.229123115539551, "genuineness": 4.545003414154053, "identity": 0.7881627678871155, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi, quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/30993bc7fe29a1ee18d1a12093686a601f260f825ac736fd2d009ea3ba8380cc.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "31598dda798d7e3f6a0b7c5854e62634792be51a81938602abe286d6de4303b1", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "973c26a6a29a1cf615f99ae8f494ba207ffcaf574d5769306f5c1bf5601621ef", "generation_seconds_batch": 22.698339018970728, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.191899061203003, "target_z": 1.1562500108079916, "ce": 6.23706579208374, "genuineness": 4.2374773025512695, "identity": 0.324369341135025, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/31598dda798d7e3f6a0b7c5854e62634792be51a81938602abe286d6de4303b1.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "31a944a836f6ccfde53d76da3dc19730a780bed2999d9a50e60f83356c6ec716", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ff0771e2a189fa3e914c70c3aa6468acda05360e9f4188558c01cba6faf42261", "generation_seconds_batch": 21.000383550999686, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.752002716064453, "target_z": 1.5624999351520503, "ce": 5.5817952156066895, "genuineness": 4.545003414154053, "identity": 0.5052119493484497, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/31a944a836f6ccfde53d76da3dc19730a780bed2999d9a50e60f83356c6ec716.mp3" }, { "id": "heldout-nl-0", "language": "nl", "sentences": [ "Je bericht kwam nadat ik vertrokken was.", "Ik las het twee keer... en legde mijn telefoon weg.", "Zeg maar wanneer we kunnen praten." ], "text": "Je bericht kwam nadat ik vertrokken was. Ik las het twee keer... en legde mijn telefoon weg. Zeg maar wanneer we kunnen praten.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-0", "key": "31a9f237fa1355cb57ad90f07cb2593ce2a1c6a7be55e42d699cc430d2f05e36", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Natural clear speech.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f1aad695600e85ec868c20cd43b6ff87f237449f65c2d15792dd13fbe76ba95e", "generation_seconds_batch": 16.65428311296273, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.31038236618042, "target_z": 1.2421873784100943, "ce": 5.8875885009765625, "genuineness": 4.369274139404297, "identity": 0.32535868883132935, "hypothesis": "Je bericht kwam nadat ik vertrokken was. Ik las het twee keer en legde mijn telefoon weg. Zeg maar wanneer we kunnen praten.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/31a9f237fa1355cb57ad90f07cb2593ce2a1c6a7be55e42d699cc430d2f05e36.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "3211d4749139cab7d514a69383a9d5aa52a5c74cc7db97f4f1bb30e143014628", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6738cbcfad59b21136f6421e5caa70d4a6a6dc624881c86986b736e95fb13053", "generation_seconds_batch": 24.495948707102798, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.437913179397583, "target_z": 0.609374972980021, "ce": 5.593709468841553, "genuineness": 3.8797430992126465, "identity": 0.42540496587753296, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás lista.", "wer": 0.06666666666666667, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3211d4749139cab7d514a69383a9d5aa52a5c74cc7db97f4f1bb30e143014628.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "32b7f30dbe5d7e580fba1b1f361dc105d930f82258e0df0326cb3a4c7d22873c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c9453e7d002a241dfa6799b275a3ca2757b07e0c017a7893a826be7bb8b27e96", "generation_seconds_batch": 23.730586632038467, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.342696189880371, "target_z": 1.2656249837880125, "ce": 5.442798614501953, "genuineness": 4.030367851257324, "identity": 0.5004388093948364, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/32b7f30dbe5d7e580fba1b1f361dc105d930f82258e0df0326cb3a4c7d22873c.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "33040c35ed37419bf93e501758c12999e6cd3ccc497a1c9d1db7d266aad0b7c6", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "478cb59444004d45f08d9f6b043edb746a6ca938b4fa3b1687793ef5c65b0947", "generation_seconds_batch": 21.8217074990971, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.019559383392334, "target_z": 1.0312499675760252, "ce": 6.268836498260498, "genuineness": 4.7834930419921875, "identity": 0.7389428615570068, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/33040c35ed37419bf93e501758c12999e6cd3ccc497a1c9d1db7d266aad0b7c6.mp3" }, { "id": "heldout-nl-2", "language": "nl", "sentences": [ "De weg gaat linksaf na de brug.", "Loop verder... tot aan het kleine houten hek.", "Ik wacht aan de andere kant." ], "text": "De weg gaat linksaf na de brug. Loop verder... tot aan het kleine houten hek. Ik wacht aan de andere kant.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-2", "key": "33f63538ac2c49676cc53462aa2024b47e58cb29b8690774a6580785ecb0e60e", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik wacht aan de andere kant.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik wacht aan de andere kant.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "84997ff7fbfe6f7f2de174397dcf3c3928e06f42ab93a83d00c51b060d4d940c", "generation_seconds_batch": 18.06750554998871, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5314061641693115, "target_z": -0.7734375189139854, "ce": 6.1020402908325195, "genuineness": 4.406930446624756, "identity": 0.7174021005630493, "hypothesis": "De weg gaat linksaf naar de brug. Loop verder tot aan het kleine houten hek. Ik wacht aan de andere kant.", "wer": 0.047619047619047616, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/33f63538ac2c49676cc53462aa2024b47e58cb29b8690774a6580785ecb0e60e.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "347abe12ae90edcc5c99833a164152ce94c5b997a97a9bbe5d2552f05bce6c31", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7ec7862520427b2d9277e707d396ecd3715fdc4dfb989259fb8265af43778d11", "generation_seconds_batch": 16.808574029942974, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4719510078430176, "target_z": 1.359375059443954, "ce": 5.919358730316162, "genuineness": 4.7458367347717285, "identity": 0.7224075794219971, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/347abe12ae90edcc5c99833a164152ce94c5b997a97a9bbe5d2552f05bce6c31.mp3" }, { "id": "heldout-en-2", "language": "en", "sentences": [ "The road turns left behind the bridge.", "Keep going... until you reach the small wooden gate.", "I will meet you on the other side." ], "text": "The road turns left behind the bridge. Keep going... until you reach the small wooden gate. I will meet you on the other side.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-2", "key": "347b084eb1c87968a3fc14b855043a2de140e81cf0874b7d65525285bfb31414", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The road turns left behind the bridge.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I will meet you on the other side.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The road turns left behind the bridge.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I will meet you on the other side.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ce057b3667da5e37a2c15eaf5bbc25926d176f4656cfd58a4f92e575ebcd2887", "generation_seconds_batch": 16.38473025301937, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.464463710784912, "target_z": -0.0966796466322817, "ce": 6.4515180587768555, "genuineness": 4.218649387359619, "identity": 0.372866153717041, "hypothesis": "The road turns left behind the bridge. Keep going until you reach the small wooden gate. I will meet you on the other side.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/347b084eb1c87968a3fc14b855043a2de140e81cf0874b7d65525285bfb31414.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "347b9fcb62c6d64c6cc488ea07a5fb22fb91a3a94a9dd95902507d674f1e5cb0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "ab2033618887d424c4de98981eeb7ebe52948ddb2afde9b7ec34aae6eea5444a", "generation_seconds_batch": 23.315708325011656, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6119768619537354, "target_z": 1.460937497298002, "ce": 5.911416053771973, "genuineness": 4.319066047668457, "identity": 0.6595666408538818, "hypothesis": "Ich habe die kleine Schachte gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0625, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/347b9fcb62c6d64c6cc488ea07a5fb22fb91a3a94a9dd95902507d674f1e5cb0.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "347d79fd791d77d5b86161983270b478951e349b7b13db1f8ba4690930c655c5", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "df8bb539d6cd14fafee6bb6a9567a460c3fa8216c776e510997ac686a4881299", "generation_seconds_batch": 22.884885416948237, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.2259368896484375, "target_z": 1.9062500972719245, "ce": 6.0146708488464355, "genuineness": 4.570107460021973, "identity": 0.7948040962219238, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/347d79fd791d77d5b86161983270b478951e349b7b13db1f8ba4690930c655c5.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "347f6c010b8787cef5afddc30513b1921d7cdf7e12eb83a6ca99fe2be518d144", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3a22acfcb3ceabab9487651913035b641848239ca4089193057c3e941ffe0942", "generation_seconds_batch": 18.084932073019445, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.07513952255249, "target_z": 1.7968749513640376, "ce": 6.0146708488464355, "genuineness": 4.532451152801514, "identity": 0.8383042216300964, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/347f6c010b8787cef5afddc30513b1921d7cdf7e12eb83a6ca99fe2be518d144.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "34cedf10f0bb39b28ccbe4e3a65e9f48a0e94c1737c75b2e1277420c5ade80c8", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d53aa1d759f80973f808e8dc4b101a54940d9888b049b43a336e086784a4156f", "generation_seconds_batch": 15.6414084819844, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.256526470184326, "target_z": 1.2031250486359624, "ce": 6.197352409362793, "genuineness": 4.030367851257324, "identity": 0.31160658597946167, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/34cedf10f0bb39b28ccbe4e3a65e9f48a0e94c1737c75b2e1277420c5ade80c8.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "35ca8e351908a3c7d88b34a96f2d35a95c2f1ea29340a4c835b3fb911331fbc9", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c548f6f49892b57569cb7e8da17634381952ca87e325488100e14c69fbeeebf2", "generation_seconds_batch": 25.705541954026558, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.847219944000244, "target_z": 0.9062500972719246, "ce": 6.24500846862793, "genuineness": 4.1998209953308105, "identity": 0.38093507289886475, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/35ca8e351908a3c7d88b34a96f2d35a95c2f1ea29340a4c835b3fb911331fbc9.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "3797882fb9d0609c746752adc85b8e56848b5eea23e7356f0a9899d7a3affb39", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "fb5c89a389f39708f8915114d7762163ad893fe29f86d07e67966f6dded86dce", "generation_seconds_batch": 23.38123934203759, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9666720628738403, "target_z": 0.2675781175695058, "ce": 6.24500846862793, "genuineness": 4.406930446624756, "identity": 0.5353415012359619, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3797882fb9d0609c746752adc85b8e56848b5eea23e7356f0a9899d7a3affb39.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "38003ad61e35ce26e0c1aff3bf1728047261f2119fe58379f3fc1bf666edbdec", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "42eb2301f6d75728f6dc0a7d6fc2b63d7648aefc843913d200b2ae7dd0974e67", "generation_seconds_batch": 21.842607220984064, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9137847423553467, "target_z": -0.4960937324370136, "ce": 6.23706579208374, "genuineness": 4.413206577301025, "identity": 0.7464679479598999, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/38003ad61e35ce26e0c1aff3bf1728047261f2119fe58379f3fc1bf666edbdec.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "38594cd3ccd5e289a54df76c95195b81f188c00b44d1a79b1eb3fddd01c1f22d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6587af66a69f92718b6353dfaa4dd0fb1f6d5e97da7d298708209298581a3e79", "generation_seconds_batch": 16.090442794957198, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8041348457336426, "target_z": 0.8749999567680335, "ce": 6.483288764953613, "genuineness": 4.557555198669434, "identity": 0.5084218382835388, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/38594cd3ccd5e289a54df76c95195b81f188c00b44d1a79b1eb3fddd01c1f22d.mp3" }, { "id": "heldout-en-1", "language": "en", "sentences": [ "The keys are on the kitchen table.", "I checked the drawer... then looked beside the window.", "Take them with you when you leave." ], "text": "The keys are on the kitchen table. I checked the drawer... then looked beside the window. Take them with you when you leave.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-1", "key": "385f71f8219765428a37b3af890a308dc263742af57be0313d35a1ab68ee777a", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The keys are on the kitchen table.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Take them with you when you leave.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The keys are on the kitchen table.\n(Natural clear speech.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Natural clear speech.) [3.2 seconds duration] Take them with you when you leave.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0e0b6cf651a7ae375a8b2ddcad37057ad26eeb10ee5f5792bb58acee41c97518", "generation_seconds_batch": 16.069452674943022, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5042645931243896, "target_z": 1.3828124918940063, "ce": 5.6413655281066895, "genuineness": 4.118232727050781, "identity": 0.3925056457519531, "hypothesis": "The keys are on the kitchen table. I checked the drawer, then looked beside the window. Take them with you when you leave.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/385f71f8219765428a37b3af890a308dc263742af57be0313d35a1ab68ee777a.mp3" }, { "id": "heldout-es-3", "language": "es", "sentences": [ "Alguien ha movido la pequeña silla azul.", "Estaba aquí... junto a la estantería alta.", "La colocaremos después de comer." ], "text": "Alguien ha movido la pequeña silla azul. Estaba aquí... junto a la estantería alta. La colocaremos después de comer.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-3", "key": "38785b4a41aef2e8ea8a3c60ebf9cbc19b5a625b61ceac90fc905118cb63f392", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La colocaremos después de comer.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Natural clear speech.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Natural clear speech.) [3.2 seconds duration] La colocaremos después de comer.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9512573045d56156804a800c6892fdd3b491bb745ce4beec2d36f898d3bad38e", "generation_seconds_batch": 22.70087477308698, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5796632766723633, "target_z": 1.4375000648479497, "ce": 5.879645347595215, "genuineness": 4.6705241203308105, "identity": 0.501914381980896, "hypothesis": "Alguien ha movido la pequeña silla azul. Estaba aquí, junto a la estantería alta. La colocaremos después de comer.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/38785b4a41aef2e8ea8a3c60ebf9cbc19b5a625b61ceac90fc905118cb63f392.mp3" }, { "id": "heldout-fr-3", "language": "fr", "sentences": [ "Quelqu’un a déplacé la petite chaise bleue.", "Elle était ici... près de la grande bibliothèque.", "Nous la remettrons après le déjeuner." ], "text": "Quelqu’un a déplacé la petite chaise bleue. Elle était ici... près de la grande bibliothèque. Nous la remettrons après le déjeuner.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-3", "key": "387a9a8d3875baa3fad7339e75812d0dfc147c093a38012f3fa061a956d0c7b8", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Natural clear speech.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Natural clear speech.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "f66d69ca3c588e4f9c5b4fa401714a7424c8c3eda152ef0c8475e9ba363844d5", "generation_seconds_batch": 24.642980829929, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4934933185577393, "target_z": 1.3749999567680335, "ce": 5.522225379943848, "genuineness": 4.419482707977295, "identity": 0.7686857581138611, "hypothesis": "Kelka the petite chaise blew. El is pretty launch bibliothek.", "wer": 0.9047619047619048, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/387a9a8d3875baa3fad7339e75812d0dfc147c093a38012f3fa061a956d0c7b8.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "38db16ca973cb85958e210b0519121dce2cb7c19399113b938f54d6873281d39", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "f1624d948de404454773bdd0b8c4c464c96c1e07b5732b48569ea46d88bd2edc", "generation_seconds_batch": 23.86717946291901, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.51503586769104, "target_z": 1.3906250270199791, "ce": 5.982900142669678, "genuineness": 4.231201171875, "identity": 0.31408631801605225, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/38db16ca973cb85958e210b0519121dce2cb7c19399113b938f54d6873281d39.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "392b91351e7a07a930f2ed0c527f544ffa854d0510b15a49c831125e61cfe718", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ffbb2bb5a96812e93519ec7071bad98cb098daa47388a69d0eca886730da2afb", "generation_seconds_batch": 25.0993486889638, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1488142013549805, "target_z": 1.1250000432319665, "ce": 6.149696350097656, "genuineness": 4.532451152801514, "identity": 0.6898053288459778, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/392b91351e7a07a930f2ed0c527f544ffa854d0510b15a49c831125e61cfe718.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "392cb4e3432b2236ab17d4fba100c9ab008a9a4ef12090498cdb5901cc963415", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d955709d8b509ef204712be064c75df9180ca2dd1641e73ef164f27db3fb4194", "generation_seconds_batch": 23.977246632101014, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.096682071685791, "target_z": 1.8125000216159832, "ce": 5.34748649597168, "genuineness": 4.406930446624756, "identity": 0.7891044616699219, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre.", "wer": 0.3684210526315789, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/392cb4e3432b2236ab17d4fba100c9ab008a9a4ef12090498cdb5901cc963415.mp3" }, { "id": "heldout-es-3", "language": "es", "sentences": [ "Alguien ha movido la pequeña silla azul.", "Estaba aquí... junto a la estantería alta.", "La colocaremos después de comer." ], "text": "Alguien ha movido la pequeña silla azul. Estaba aquí... junto a la estantería alta. La colocaremos después de comer.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-3", "key": "3969aeec1df7d7e49707083840c8e55b8cb86612d08ebbb3bfa3ec8a524a8151", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La colocaremos después de comer.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Natural clear speech.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Natural clear speech.) [3.2 seconds duration] La colocaremos después de comer.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "78befb1ca01ba412f3821bd3e4d4d2fb5b170cf588469786c4ba0d8a725ff19d", "generation_seconds_batch": 18.06453359290026, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.245755195617676, "target_z": 1.1953125135099896, "ce": 6.0305562019348145, "genuineness": 4.632867813110352, "identity": 0.3920097053050995, "hypothesis": "Alguien ha movido la pequeña silla azul. Estaba aquí, junto a la estantería alta. La colocaremos después de comer.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3969aeec1df7d7e49707083840c8e55b8cb86612d08ebbb3bfa3ec8a524a8151.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "3a0528c94e7ad3492d93939ba2a798b5fb45d39569318403d7de732d4984fc94", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "286c9f704c4c5d76b255e5eabee0fd1bb69963b8bdebb2599cfe43c776ee58e0", "generation_seconds_batch": 23.269659119076096, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.31038236618042, "target_z": 1.2421873784100943, "ce": 6.062326908111572, "genuineness": 4.457139015197754, "identity": 0.5956578850746155, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3a0528c94e7ad3492d93939ba2a798b5fb45d39569318403d7de732d4984fc94.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "3b5d41371e596425dd26be52c6cb11e6177a3501e4d5e1decf0c71361ce03dde", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "bd4fdadebe8f6b8f9902e4390f3f1d385000722630c67e4db06bcaaf249be5fa", "generation_seconds_batch": 22.80958472704515, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.07513952255249, "target_z": 1.7968749513640376, "ce": 5.756534099578857, "genuineness": 4.268857479095459, "identity": 0.31827694177627563, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3b5d41371e596425dd26be52c6cb11e6177a3501e4d5e1decf0c71361ce03dde.mp3" }, { "id": "heldout-es-1", "language": "es", "sentences": [ "Las llaves están en la mesa de la cocina.", "Miré en el cajón... y luego junto a la ventana.", "Llévatelas cuando te vayas." ], "text": "Las llaves están en la mesa de la cocina. Miré en el cajón... y luego junto a la ventana. Llévatelas cuando te vayas.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-1", "key": "3b8a142d4cb45308366d93133e200680a808e1b20eada2a0883d4b54d64d508b", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Llévatelas cuando te vayas.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Natural clear speech.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Natural clear speech.) [3.2 seconds duration] Llévatelas cuando te vayas.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "bee8aaad4d1873d25f54fe0ce85de7cd358b7f2520a2ce6b604c5e5e221d7998", "generation_seconds_batch": 17.253805133979768, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5365781784057617, "target_z": 1.4062499243440587, "ce": 5.6334228515625, "genuineness": 4.3504462242126465, "identity": 0.5492691993713379, "hypothesis": "Las llaves están en la mesa de la cocina. Mira en el cajón y luego, junto a la ventana, llévatelas cuando te vayas.", "wer": 0.043478260869565216, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3b8a142d4cb45308366d93133e200680a808e1b20eada2a0883d4b54d64d508b.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "3baeae24db4b5bcded3a2fba726e1fcbc8f8ff74a02c80cd0a893a7d363b2ce1", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "adafdce9f125d49e4af2936854b5370e434da24e4cb1cba7e2379f25153724fe", "generation_seconds_batch": 24.505743153975345, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.441361427307129, "target_z": 2.062500108079916, "ce": 5.589737892150879, "genuineness": 4.482243061065674, "identity": 0.6671794652938843, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3baeae24db4b5bcded3a2fba726e1fcbc8f8ff74a02c80cd0a893a7d363b2ce1.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "3bc9d425eac7a43c2d8263ad223296e55d629196f3cef05f1c0a3533a6f62b35", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "675621fe7d66515b8770dab0f5e0ee6f1f2175f7552cb4e669729a86db927753", "generation_seconds_batch": 25.344684287905693, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4719510078430176, "target_z": 1.359375059443954, "ce": 6.125868320465088, "genuineness": 4.720732688903809, "identity": 0.6314623355865479, "hypothesis": "Te encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.058823529411764705, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3bc9d425eac7a43c2d8263ad223296e55d629196f3cef05f1c0a3533a6f62b35.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "3bcdfdfa97b8eb852613c400793d6bd3ffae1775deb08da13f9bc337c8e8578c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b77664548c6595fd274220c61b0969c2f43b080211e6c7bcabb6fef9bce2dae8", "generation_seconds_batch": 17.018014376983047, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3857812881469727, "target_z": 1.2968751242919037, "ce": 6.213237762451172, "genuineness": 4.582659721374512, "identity": 0.5584850311279297, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3bcdfdfa97b8eb852613c400793d6bd3ffae1775deb08da13f9bc337c8e8578c.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "3c0a029fe9b9ce78faf5df7c4caf863eb9f2877acdc3983638d0e01428bf37a9", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1b27ed7f5662977a92de867e4cb4537f7b20c4dd86f26e70b83c76b930ce53bf", "generation_seconds_batch": 23.33019120595418, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3534674644470215, "target_z": 1.2734375189139853, "ce": 5.8875885009765625, "genuineness": 4.444586753845215, "identity": 0.6044279336929321, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3c0a029fe9b9ce78faf5df7c4caf863eb9f2877acdc3983638d0e01428bf37a9.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "3c0a540741e4fc72588ba72ec3c1ed807c4f0171a0dab807ed633cefc12f2d89", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "87e3c823c3d8fc5b26a17886084ddfcdd1b7643e217100878698747a437ee6d6", "generation_seconds_batch": 24.883491652086377, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.965703248977661, "target_z": 0.9921874648740272, "ce": 6.125868320465088, "genuineness": 4.262581825256348, "identity": 0.27694571018218994, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3c0a540741e4fc72588ba72ec3c1ed807c4f0171a0dab807ed633cefc12f2d89.mp3" }, { "id": "heldout-nl-1", "language": "nl", "sentences": [ "De sleutels liggen op de keukentafel.", "Ik keek in de la... en daarna naast het raam.", "Neem ze mee als je weggaat." ], "text": "De sleutels liggen op de keukentafel. Ik keek in de la... en daarna naast het raam. Neem ze mee als je weggaat.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-1", "key": "3c590f9302048ad206a88d3c2440f1028cc9eea1f1662250baf365be21a6ebfa", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Neem ze mee als je weggaat.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Neem ze mee als je weggaat.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6bf07a3f1be35b829224c8320cfbb428706ad8d585430f146b249cc7b6c05c7b", "generation_seconds_batch": 17.59364196297247, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0888171195983887, "target_z": -0.3691406418874869, "ce": 6.300607204437256, "genuineness": 4.293961524963379, "identity": 0.5061303377151489, "hypothesis": "De sleutels liggen op de keukentafel. Ik keek in de la en daarna naast het raam. Neem ze mee als je weggaat.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3c590f9302048ad206a88d3c2440f1028cc9eea1f1662250baf365be21a6ebfa.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "3c5d125c29bf4a62d3c974f5baec4ee0762392a848a65786d5fab1071eb7aebe", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b3a2b84f794aafd22c63336dc6d34b51e3ccead04bbe9015ee35fb393b477d53", "generation_seconds_batch": 24.883491652086377, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6802659034729004, "target_z": 0.7851563216029445, "ce": 6.372091293334961, "genuineness": 4.3127899169921875, "identity": 0.46994492411613464, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3c5d125c29bf4a62d3c974f5baec4ee0762392a848a65786d5fab1071eb7aebe.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "3c9a586d161d69c3b264d99b0c420537bd495c18e93ff598b97e99298ff99964", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1e5776004bc05fb6665b0e2254ce8d96d4961da58a2ebca199e6f639404ad713", "generation_seconds_batch": 21.000383550999686, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5258069038391113, "target_z": 1.398437389218086, "ce": 5.756534099578857, "genuineness": 4.595211505889893, "identity": 0.6362311840057373, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3c9a586d161d69c3b264d99b0c420537bd495c18e93ff598b97e99298ff99964.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "3cf0d8c90ecc78ec17cd4ed8eedd474b5f3f663888f1ae47777f6058cd7e2fa0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ad6cf43e7183f72d5be6c6298bfce5b700a96a259e4e6cfccb147b130e5fbe69", "generation_seconds_batch": 24.232628850033507, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.1720805168151855, "target_z": 1.8671874216420608, "ce": 5.959072113037109, "genuineness": 4.632867813110352, "identity": 0.8320513963699341, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3cf0d8c90ecc78ec17cd4ed8eedd474b5f3f663888f1ae47777f6058cd7e2fa0.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "3cf4f9875d49552a12bb3e93a25e8553240e235b3e9430a6e0feafce00bff83e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "b81f1b1ac61960e5a26db88594e7f84079e068ff2756affbc8207833fc4d5768", "generation_seconds_batch": 23.943483604001813, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.450408458709717, "target_z": 1.3437499891920084, "ce": 5.935244083404541, "genuineness": 4.086852550506592, "identity": 0.5387601852416992, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinführen.", "wer": 0.0625, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3cf4f9875d49552a12bb3e93a25e8553240e235b3e9430a6e0feafce00bff83e.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "3d069a33d28894d8d33bc33f5e89d21f30eb69040c295a2d49bf61dca6c5983f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "23318ff6af5f41c116718d1012f45f9e96be2eac16548f8bb0d4c2fcde9e1b3e", "generation_seconds_batch": 22.9791196139995, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4396371841430664, "target_z": 1.3359374540660356, "ce": 5.6532793045043945, "genuineness": 4.457139015197754, "identity": 0.39877355098724365, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3d069a33d28894d8d33bc33f5e89d21f30eb69040c295a2d49bf61dca6c5983f.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "3e3159cfe250a88f1063cf2a9fa41cb3fd18cdc4bf26602f543231888dd426b2", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "379b22e90274b53335de22a9887b45ab42095a310e0f38b4b19d2284b186bef1", "generation_seconds_batch": 22.486650948994793, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7304604053497314, "target_z": 1.5468750378279708, "ce": 5.677107334136963, "genuineness": 4.7332844734191895, "identity": 0.6039533615112305, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3e3159cfe250a88f1063cf2a9fa41cb3fd18cdc4bf26602f543231888dd426b2.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "3e4fb780b7e8b60a7add2034bddbbfaa7d7a58e45ac6d21dce9ddac515dc3f37", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1fbd62bd4023c39e8094925316d10da07d233cc1c3b31430323c2ca970838233", "generation_seconds_batch": 24.64519719593227, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7772068977355957, "target_z": 0.8554687918809675, "ce": 6.316492557525635, "genuineness": 4.494794845581055, "identity": 0.3921058773994446, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3e4fb780b7e8b60a7add2034bddbbfaa7d7a58e45ac6d21dce9ddac515dc3f37.mp3" }, { "id": "heldout-en-3", "language": "en", "sentences": [ "Someone moved the little blue chair.", "It used to stand here... beside the tall bookshelf.", "We can put it back after lunch." ], "text": "Someone moved the little blue chair. It used to stand here... beside the tall bookshelf. We can put it back after lunch.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-3", "key": "3ed09db510863fbef230e50daa31a5446e547eea36683ee58eb52d981689e863", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Someone moved the little blue chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can put it back after lunch.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Someone moved the little blue chair.\n(Natural clear speech.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Natural clear speech.) [3.2 seconds duration] We can put it back after lunch.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c1dd0b37f74c48996ebd687674f9521f40df172b28df2db8c996da6bac16ce03", "generation_seconds_batch": 27.698526351014152, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4719510078430176, "target_z": 1.359375059443954, "ce": 6.0305562019348145, "genuineness": 4.319066047668457, "identity": 0.3409859240055084, "hypothesis": "Someone moved the little blue chair. It used to stand here beside the tall bookshelf. We can put it back after lunch.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3ed09db510863fbef230e50daa31a5446e547eea36683ee58eb52d981689e863.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "3ed7a173e1fb2a0351198d5acf454039071221d2f54b6cb3d43d45ca8a13b20e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9079e9963a2e7c687b94e43c88ca9d5e132cdd75e068085fa7351355473ce971", "generation_seconds_batch": 27.454596178024076, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1165003776550293, "target_z": 1.1015624378540483, "ce": 5.6532793045043945, "genuineness": 4.218649387359619, "identity": 0.1696174144744873, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3ed7a173e1fb2a0351198d5acf454039071221d2f54b6cb3d43d45ca8a13b20e.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "3f4224f754ca5e7a727dc890fc4b460e61ee9ea916baea820ba25bfe130782b8", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "cf3a4a28aa3909ff2cc9b000fb33e5d1f51ec0175546824aa554f5dfa06b2d18", "generation_seconds_batch": 24.1761207790114, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6442904472351074, "target_z": 1.4843749297480544, "ce": 5.796247482299805, "genuineness": 4.18099308013916, "identity": 0.364524781703949, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3f4224f754ca5e7a727dc890fc4b460e61ee9ea916baea820ba25bfe130782b8.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "3f44c25b28a306aa6dc8db29d256bcca2c165638871f990070ad989128fb9200", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "41fe96c044201480925a4c2df697ad2785a908246cd745fca32f064a2ece36bf", "generation_seconds_batch": 24.99255053792149, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2134413719177246, "target_z": 1.1718749081320712, "ce": 6.086154937744141, "genuineness": 4.218649387359619, "identity": 0.8387141227722168, "hypothesis": "La porte est encore ouverte, j'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3f44c25b28a306aa6dc8db29d256bcca2c165638871f990070ad989128fb9200.mp3" }, { "id": "heldout-nl-4", "language": "nl", "sentences": [ "Ik schreef het nummer op de envelop.", "Lees het goed... en controleer de laatste twee cijfers.", "Bel me wanneer je even tijd hebt." ], "text": "Ik schreef het nummer op de envelop. Lees het goed... en controleer de laatste twee cijfers. Bel me wanneer je even tijd hebt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-4", "key": "3f70f1ad81587f7d446cb0357254ba3316b24269fa55ebd1b9e6a4775f4a5140", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Natural clear speech.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Natural clear speech.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "dbec689cb21d8261f11419bfb316bcce46f3156e306cd77f434bdbb1fc9bf425", "generation_seconds_batch": 24.535028933081776, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.901075839996338, "target_z": 0.9453124270460566, "ce": 5.732706069946289, "genuineness": 4.18099308013916, "identity": 0.5029436945915222, "hypothesis": "Ik schreef het nummer op de envelop. Lees het goed en controleer de laatste twee cijfers. Bel me wanneer je even tijd hebt.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3f70f1ad81587f7d446cb0357254ba3316b24269fa55ebd1b9e6a4775f4a5140.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "3f8a5db414575edbb92e7921e7e064933e221bfe4bcae67a634b798af1a6ae83", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "54a8fb8b36ceaa5d2b1fbe476e37c6986ddeb041e8598a667f1350047ce84d5e", "generation_seconds_batch": 24.547843346023, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.441361427307129, "target_z": 2.062500108079916, "ce": 5.609594821929932, "genuineness": 4.469690799713135, "identity": 0.865064799785614, "hypothesis": "J'ai trouvé la petite boite. Elle était près de la vieille fenêtre.", "wer": 0.42105263157894735, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/3f8a5db414575edbb92e7921e7e064933e221bfe4bcae67a634b798af1a6ae83.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "4044bb63ddf19726ce015a3713e1f666ff514c718eaf87b4e6ee379d39210c6a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3ba60aab571c0805cffb5b8cce3e9a4c5e0d8acc10211cc884a7e57a41207913", "generation_seconds_batch": 17.02645694790408, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.204394340515137, "target_z": 1.8906250270199791, "ce": 5.831989288330078, "genuineness": 4.557555198669434, "identity": 0.7198976874351501, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4044bb63ddf19726ce015a3713e1f666ff514c718eaf87b4e6ee379d39210c6a.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "40aa50a3c6316a9eac3cd300f0b95380014183de8c3a8b6a7b974fe0abff7c9d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3e7bd18c3b6a99d122d0b423e8bd554c5b623538ffa10a838e5f196e4a30449f", "generation_seconds_batch": 22.61160150100477, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.678328514099121, "target_z": 2.2343751891398536, "ce": 5.550024509429932, "genuineness": 4.582659721374512, "identity": 0.6723483800888062, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/40aa50a3c6316a9eac3cd300f0b95380014183de8c3a8b6a7b974fe0abff7c9d.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "40c115af0572d21260ccafaa07af0c141ccf94700c7245889531f1ae7ffa10d3", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f5b11ed602f238f96d76c9cc0ba31b457917b303af1b9291522c0c0fda12f2e3", "generation_seconds_batch": 23.474032889003865, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0841867923736572, "target_z": 1.078125005403996, "ce": 6.149696350097656, "genuineness": 4.337893962860107, "identity": 0.6189334392547607, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/40c115af0572d21260ccafaa07af0c141ccf94700c7245889531f1ae7ffa10d3.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "41512750e824a9b8485d8f2508fd6a74fa12304b05792b1dec540ea2ca56c77a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6e8449093277a5129a9d4a8bef4905ee48978e6649412124ff2268299d350222", "generation_seconds_batch": 25.484242056962103, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.601205587387085, "target_z": 1.4531249621720292, "ce": 5.839932441711426, "genuineness": 4.369274139404297, "identity": 0.7417739629745483, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/41512750e824a9b8485d8f2508fd6a74fa12304b05792b1dec540ea2ca56c77a.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "415fb7f7042c36347696d2c743ce9313a92598563ea7b6781b61b10a82becc02", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "c524190604058f3d98e660c8e6d4b901a0213f95509f4b418687cdc2ec0dbcec", "generation_seconds_batch": 23.686055146041326, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7627739906311035, "target_z": 1.5703124702780231, "ce": 5.982900142669678, "genuineness": 4.381826400756836, "identity": 0.46634548902511597, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinführen.", "wer": 0.0625, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/415fb7f7042c36347696d2c743ce9313a92598563ea7b6781b61b10a82becc02.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "4179ab7796955119674c1651507d360ea93aeda4d76b35a99244d28397933f64", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9a0c8182b2f6c5ffa54b0936c5c01243f61d5e74f1a09c31736832e1ba4af4ba", "generation_seconds_batch": 17.92267002305016, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.723350763320923, "target_z": 0.8164062891789696, "ce": 6.316492557525635, "genuineness": 4.337893962860107, "identity": 0.7144827842712402, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4179ab7796955119674c1651507d360ea93aeda4d76b35a99244d28397933f64.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "4252d22ee2fb4e0098db18e90270599876530776cecb31448cfc9e86198c3205", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7b8e397f89997caf17c9838ce6a38ae1937796dca8ab99501fa20528d940d552", "generation_seconds_batch": 23.905337012954988, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.202670097351074, "target_z": 1.1640623730060984, "ce": 5.863759994506836, "genuineness": 4.419482707977295, "identity": 0.5350692272186279, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4252d22ee2fb4e0098db18e90270599876530776cecb31448cfc9e86198c3205.mp3" }, { "id": "heldout-fr-0", "language": "fr", "sentences": [ "Ton message est arrivé après mon départ.", "Je l’ai relu... puis j’ai rangé mon téléphone.", "Dis-moi quand nous pourrons parler." ], "text": "Ton message est arrivé après mon départ. Je l’ai relu... puis j’ai rangé mon téléphone. Dis-moi quand nous pourrons parler.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-0", "key": "42809c94a59170eecb3859c3b75e2930d9b55d07850995f3a682e2343218b09b", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Natural clear speech.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c12e518370532724cf7d82377a04d3477b519cdc7e1446910630c0fb5ee7a818", "generation_seconds_batch": 15.552651091013104, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9566562175750732, "target_z": 1.710937583761935, "ce": 5.534139156341553, "genuineness": 4.419482707977295, "identity": 0.7408183813095093, "hypothesis": "Two message is arrived after my depart. Je l'ai relu, puis j'ai rangé mon téléphone. Dis-moi quand nous pourrons partir.", "wer": 0.3333333333333333, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/42809c94a59170eecb3859c3b75e2930d9b55d07850995f3a682e2343218b09b.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "42b7dd328b45e4a8313a046d633aab100d391179e88f6af87b0512623fa16df8", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1b6663935fca2daa41752268e1418920b459b9048ca724c5daa2114f0d1f9ff2", "generation_seconds_batch": 15.610294860904105, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2709591388702393, "target_z": 0.48828119731104086, "ce": 6.38797664642334, "genuineness": 4.400654315948486, "identity": 0.30853700637817383, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/42b7dd328b45e4a8313a046d633aab100d391179e88f6af87b0512623fa16df8.mp3" }, { "id": "heldout-nl-2", "language": "nl", "sentences": [ "De weg gaat linksaf na de brug.", "Loop verder... tot aan het kleine houten hek.", "Ik wacht aan de andere kant." ], "text": "De weg gaat linksaf na de brug. Loop verder... tot aan het kleine houten hek. Ik wacht aan de andere kant.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-2", "key": "434f953004ee3838b31e4f505ba0cb9202b34ab37936979fff50edb64a6a996f", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik wacht aan de andere kant.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik wacht aan de andere kant.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "cb476ac7c1dbf61a3d0ae9363ac2dff34449333788f6e242aedda5c7516f021d", "generation_seconds_batch": 18.06750554998871, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5906479358673096, "target_z": -0.730468748649001, "ce": 6.229123115539551, "genuineness": 4.557555198669434, "identity": 0.7899413108825684, "hypothesis": "De weg gaat linksaf naar de brug. Loop verder tot aan het kleine Houten hek. Ik wacht aan de andere kant.", "wer": 0.047619047619047616, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/434f953004ee3838b31e4f505ba0cb9202b34ab37936979fff50edb64a6a996f.mp3" }, { "id": "heldout-fr-3", "language": "fr", "sentences": [ "Quelqu’un a déplacé la petite chaise bleue.", "Elle était ici... près de la grande bibliothèque.", "Nous la remettrons après le déjeuner." ], "text": "Quelqu’un a déplacé la petite chaise bleue. Elle était ici... près de la grande bibliothèque. Nous la remettrons après le déjeuner.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-3", "key": "437548edcff6255ddcdeee9fe3b1ac505064529078a73325d6c53a87849a6f90", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Natural clear speech.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Natural clear speech.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "7f37d1e8c0fa133aaaa049b2db4b4e1106042f926f51bfe13fd2387fe472cd7d", "generation_seconds_batch": 15.301845128997229, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3534674644470215, "target_z": 1.2734375189139853, "ce": 5.704906940460205, "genuineness": 4.545003414154053, "identity": 0.7612012624740601, "hypothesis": "Quelqu'un a déplacé la petite chaise bleue. Elle était ici, près de la grande bibliothèque. Nous la remettrons après le déjeuner.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/437548edcff6255ddcdeee9fe3b1ac505064529078a73325d6c53a87849a6f90.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "43aa69d351045b8ec259b567c57f4957966468aeaf2a8725838237838186a664", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f629ecbdc6977df2f485f4aaa31b0792d34c14b00f71ff423c40c43abad7e700", "generation_seconds_batch": 23.685131667996757, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1272716522216797, "target_z": 1.1093749729800209, "ce": 6.252951145172119, "genuineness": 4.7332844734191895, "identity": 0.7960093021392822, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/43aa69d351045b8ec259b567c57f4957966468aeaf2a8725838237838186a664.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "43ac4ea02f16aa9688f1d45dcf85f1e03a952dfaeea10f4b84d4d04740cba020", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "34188ee40ea8d2256d52c7ed0d1fe7af0a049fc33b720eb2c768546afd925f92", "generation_seconds_batch": 22.894514483981766, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6317954063415527, "target_z": 0.7500000864639329, "ce": 5.95112943649292, "genuineness": 4.231201171875, "identity": 0.36373603343963623, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/43ac4ea02f16aa9688f1d45dcf85f1e03a952dfaeea10f4b84d4d04740cba020.mp3" }, { "id": "heldout-es-3", "language": "es", "sentences": [ "Alguien ha movido la pequeña silla azul.", "Estaba aquí... junto a la estantería alta.", "La colocaremos después de comer." ], "text": "Alguien ha movido la pequeña silla azul. Estaba aquí... junto a la estantería alta. La colocaremos después de comer.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-3", "key": "43d771a5c055e2aea1186dc99f4cdef2b4c6bc8c6b2779ad35a29ca11c75a91c", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La colocaremos después de comer.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La colocaremos después de comer.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "27cb3b367f3d577f2b2e4b7fd874f8b045a525c228a63ccf6dd4010af45ece32", "generation_seconds_batch": 17.430664474959485, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.592371940612793, "target_z": -0.003906267562986381, "ce": 6.268836498260498, "genuineness": 4.645420074462891, "identity": 0.5142759084701538, "hypothesis": "Alguien ha movido la pequeña silla azul. Estaba aquí, junto a la estantería alta. La colocaremos después de comer.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/43d771a5c055e2aea1186dc99f4cdef2b4c6bc8c6b2779ad35a29ca11c75a91c.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "44062dc80a6cd55938f9418dc30d032fe4603c9167e8fb03fbc4e3ccd4031484", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "694aacf2e6ab13720f1ac2097c999f231310fa85e37a2baa5e3cd7a17b7324f8", "generation_seconds_batch": 23.685131667996757, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6210241317749023, "target_z": 0.7421875513379602, "ce": 6.086154937744141, "genuineness": 4.65797233581543, "identity": 0.784175455570221, "hypothesis": "J'ai trouvé la petite j'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.21052631578947367, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/44062dc80a6cd55938f9418dc30d032fe4603c9167e8fb03fbc4e3ccd4031484.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "4422d57a6ce0ead2a9bb260a02f96ca417bdffe68a630b62a1ab5d6cef2543a9", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a8e75f91871279c0be43afa4514e5c7e9c01295b059b703d06f8201a19d1cd43", "generation_seconds_batch": 25.293777625076473, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.838172674179077, "target_z": 1.6250000432319665, "ce": 5.736677646636963, "genuineness": 4.557555198669434, "identity": 0.6469734311103821, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4422d57a6ce0ead2a9bb260a02f96ca417bdffe68a630b62a1ab5d6cef2543a9.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "44fb9043f0fe29c0edaab8037103688c785e898cc1c949c350cdd0815123f2ca", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0ad2e580b344dcfbc1ef71ffb5ab8c7e8fc923c888306addcf49ebbb3255922c", "generation_seconds_batch": 25.912818546989, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9087767601013184, "target_z": 0.22558595742723456, "ce": 6.260893821716309, "genuineness": 4.281409740447998, "identity": 0.6731224060058594, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/44fb9043f0fe29c0edaab8037103688c785e898cc1c949c350cdd0815123f2ca.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "450884e4c965ef04afe68d359a10824680bcdbfd76aff6b34fe9482e26f9872e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "90b65987a8d57098c30da494188faaf4a649b1bff5b6a2ed5af06775e51baced", "generation_seconds_batch": 26.81906083296053, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2672977447509766, "target_z": 1.210937583761935, "ce": 6.260893821716309, "genuineness": 4.68307638168335, "identity": 0.6908239126205444, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/450884e4c965ef04afe68d359a10824680bcdbfd76aff6b34fe9482e26f9872e.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "451a4a84260f1dafd5de6847d025a683ef3362a1f412278e4bb7e476b7158d8e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "546957feabebda6e57d186c154dfab0a20e9e11f6a6184348d8c9d295fe7212c", "generation_seconds_batch": 22.80958472704515, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.601205587387085, "target_z": 1.4531249621720292, "ce": 5.863759994506836, "genuineness": 4.042920112609863, "identity": 0.3120386004447937, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/451a4a84260f1dafd5de6847d025a683ef3362a1f412278e4bb7e476b7158d8e.mp3" }, { "id": "heldout-de-4", "language": "de", "sentences": [ "Ich schrieb die Nummer auf den Umschlag.", "Lies sie genau... und prüfe die letzten beiden Ziffern.", "Ruf an, wenn du Zeit hast." ], "text": "Ich schrieb die Nummer auf den Umschlag. Lies sie genau... und prüfe die letzten beiden Ziffern. Ruf an, wenn du Zeit hast.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-4", "key": "45f9ec8497c8691f9e18d794cf98f47c522bfab041b36df5accabf5c5c456412", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "768bfd951ed7b0c5e82456d92b523a921c774cdb567dd7366ee6960bffe449f9", "generation_seconds_batch": 17.61561208299827, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9245559573173523, "target_z": -0.4882812405430073, "ce": 6.340320587158203, "genuineness": 4.193545341491699, "identity": 0.5766547322273254, "hypothesis": "Ich schrieb die Nummer auf den Umschlag, lies sie genau und prüfe die letzten beiden Ziffern. Ruf an, wenn du Zeit hast.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/45f9ec8497c8691f9e18d794cf98f47c522bfab041b36df5accabf5c5c456412.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "4617ce0f0b415ed038672394410a14b3b65906d5c6dd894349e0a6bb72078149", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "97fd7b839647bcb483525ef6e61e4dc25cd6f8ad93edc1ffbd1b911cfae68ec2", "generation_seconds_batch": 22.28326332499273, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6596921682357788, "target_z": 0.04492190404647747, "ce": 6.316492557525635, "genuineness": 4.319066047668457, "identity": 0.5607140064239502, "hypothesis": "ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen daar samen in kijken.", "wer": 0.05555555555555555, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4617ce0f0b415ed038672394410a14b3b65906d5c6dd894349e0a6bb72078149.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "462c00d2b4d4c93d841308f0d56766687347888500297fdd408b5ee1fc2cfa9f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "798d7b5aebe1d72e0c41ad6af96488170deb3bf9c923620249b037f3cad2ba30", "generation_seconds_batch": 21.956297968048602, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.527531147003174, "target_z": 2.1250000432319665, "ce": 5.8875885009765625, "genuineness": 4.482243061065674, "identity": 0.7617034316062927, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/462c00d2b4d4c93d841308f0d56766687347888500297fdd408b5ee1fc2cfa9f.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "4632b95e5d374de631bc0d9d68388054b6fc8604719840e334697344b683915d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "eab6ec8794be366debe6778e02333295e694dcdf367ab75f16d4e7e593bfc181", "generation_seconds_batch": 24.058624504017644, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5940959453582764, "target_z": 0.7226562135230283, "ce": 6.23706579208374, "genuineness": 4.7081804275512695, "identity": 0.6789138317108154, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4632b95e5d374de631bc0d9d68388054b6fc8604719840e334697344b683915d.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "469fd6246cc1a4f87d8eddfb691ec531d3413ff72d8371008af9671738939ac0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "00d751227f14cc47d9549d591211d1579d66ce39828d64749e5b5fb184a8a0b9", "generation_seconds_batch": 24.1761207790114, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.51503586769104, "target_z": 1.3906250270199791, "ce": 5.792276382446289, "genuineness": 4.268857479095459, "identity": 0.4111337959766388, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/469fd6246cc1a4f87d8eddfb691ec531d3413ff72d8371008af9671738939ac0.mp3" }, { "id": "heldout-de-1", "language": "de", "sentences": [ "Die Schlüssel liegen auf dem Küchentisch.", "Ich suchte in der Schublade... dann neben dem Fenster.", "Nimm sie mit, wenn du gehst." ], "text": "Die Schlüssel liegen auf dem Küchentisch. Ich suchte in der Schublade... dann neben dem Fenster. Nimm sie mit, wenn du gehst.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-1", "key": "47082ce3686d546e4aee934fe6840fe519e38f970b645775f4c0f8de90e751e4", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Natural clear speech.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Natural clear speech.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "895c5eeb6e2cf0f9a5988a3ed8c45e63c0451cdba8398ea2c2b512c43577437c", "generation_seconds_batch": 16.35157260799315, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6227481365203857, "target_z": 1.4687500324239748, "ce": 5.6334228515625, "genuineness": 4.432034492492676, "identity": 0.35619056224823, "hypothesis": "Die Schlösser liegen auf dem Küchentisch. Ich suchte in der Schublade, dann neben dem Fenster. Nimm sie mit, wenn du gehst.", "wer": 0.047619047619047616, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/47082ce3686d546e4aee934fe6840fe519e38f970b645775f4c0f8de90e751e4.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "4736a8cae6cf3b8c2102b1765cb83386fb8e317257aa26859dc5c7639b28389e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "504277698d15f3335d2566487087a6c2ba928bf689d94eaefe86528bc57d9e53", "generation_seconds_batch": 22.008515903027728, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1811277866363525, "target_z": 1.1484374756820188, "ce": 6.070269584655762, "genuineness": 4.482243061065674, "identity": 0.6364454030990601, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4736a8cae6cf3b8c2102b1765cb83386fb8e317257aa26859dc5c7639b28389e.mp3" }, { "id": "heldout-fr-3", "language": "fr", "sentences": [ "Quelqu’un a déplacé la petite chaise bleue.", "Elle était ici... près de la grande bibliothèque.", "Nous la remettrons après le déjeuner." ], "text": "Quelqu’un a déplacé la petite chaise bleue. Elle était ici... près de la grande bibliothèque. Nous la remettrons après le déjeuner.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-3", "key": "47d00b404f769e82194cfb39f2be7a6bd3cc5bfaac4f38356b0fada2abdb9f4d", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6d5210be00a81a20e6997ad1562c72ae62a92f7c7cac2cbaf2c0a480397f485f", "generation_seconds_batch": 17.520918608061038, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.235952854156494, "target_z": 0.46289071754342825, "ce": 6.229123115539551, "genuineness": 4.406930446624756, "identity": 0.616790771484375, "hypothesis": "Quelqu'un a déplacé la petite chaise bleue. Elle était ici, près de la grande bibliothèque. Nous la remettrons après le déjeuner.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/47d00b404f769e82194cfb39f2be7a6bd3cc5bfaac4f38356b0fada2abdb9f4d.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "4812a273a12221ab552dffebb1f2c1a02be52252bc3474a6ce2f536715b48485", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "b62de5ea3a19486cf845ed416c94e0c7ed048ed74155db9e848f15cfe0af2e60", "generation_seconds_batch": 25.24895870196633, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.805858850479126, "target_z": 1.6015624378540483, "ce": 6.125868320465088, "genuineness": 4.582659721374512, "identity": 0.8350332975387573, "hypothesis": "J'ai trouvé la petite boîte, elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4812a273a12221ab552dffebb1f2c1a02be52252bc3474a6ce2f536715b48485.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "4868f56c49ccc65ba0a1a85ae6e1296505f92e6bd4246dd33b2e2c49b26c222b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "077a661f13a276555f3e429b75385d075789e01cff161820b38202b03ba81168", "generation_seconds_batch": 24.05071882903576, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.224212646484375, "target_z": 1.179687443258044, "ce": 6.24500846862793, "genuineness": 3.9927117824554443, "identity": 0.3987366557121277, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4868f56c49ccc65ba0a1a85ae6e1296505f92e6bd4246dd33b2e2c49b26c222b.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "48a53ee354d7c33e373a86acaae737237c45ac72a3da2f61170911fd7eebc05f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "16f105478bc465a47e2bc09398129958f516b2b6242887c3d6a78cac1729e053", "generation_seconds_batch": 23.192518897005357, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.107453346252441, "target_z": 1.820312556741956, "ce": 5.990842819213867, "genuineness": 4.419482707977295, "identity": 0.6759446859359741, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/48a53ee354d7c33e373a86acaae737237c45ac72a3da2f61170911fd7eebc05f.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "48f7500b445054340864d23cf5eb8064323ca18c3623b477e497aff973e999b0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f79ce6ad71a880487997fe1d2f99187b927b4b38c04dd163fa2720a6e18e3c76", "generation_seconds_batch": 24.634708809899166, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5258069038391113, "target_z": 1.398437389218086, "ce": 5.95112943649292, "genuineness": 4.243753433227539, "identity": 0.3235875964164734, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/48f7500b445054340864d23cf5eb8064323ca18c3623b477e497aff973e999b0.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "499e51c6541b6e43b1954d9636605e50162248f53daa5bfc1221eda08b54360d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1d5433d5bad7d22fa7f272d757bf1f7e367c6d06c605aabd978bb28aa3b55379", "generation_seconds_batch": 25.95690680795815, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.691037178039551, "target_z": 0.7929688567289173, "ce": 5.8875885009765625, "genuineness": 4.105680465698242, "identity": 0.5790168046951294, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás lista.", "wer": 0.06666666666666667, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/499e51c6541b6e43b1954d9636605e50162248f53daa5bfc1221eda08b54360d.mp3" }, { "id": "heldout-es-0", "language": "es", "sentences": [ "Tu mensaje llegó después de que me fuera.", "Lo leí dos veces... y guardé el teléfono.", "Dime cuándo podemos hablar." ], "text": "Tu mensaje llegó después de que me fuera. Lo leí dos veces... y guardé el teléfono. Dime cuándo podemos hablar.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-0", "key": "49d468ae6540cc7c1e5b0e6883c25e2b29599458c9de0a5398000eaca62677ab", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo podemos hablar.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo podemos hablar.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ca6972ca88534939bbaa3a1d667e7fd11c03195c3259caa1a7f6df7634415e67", "generation_seconds_batch": 17.511108755948953, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5062021017074585, "target_z": -0.06640628917896962, "ce": 6.213237762451172, "genuineness": 4.042920112609863, "identity": 0.5507676601409912, "hypothesis": "Tu mensaje llegó después de que me fuera. Lo leí dos veces y guardé el teléfono. Dime cuándo podemos hablar.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/49d468ae6540cc7c1e5b0e6883c25e2b29599458c9de0a5398000eaca62677ab.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "4b5c8df10a9ce7d470891db09c58b0de15ade72b7302ff37ea7c19bc6d28daf6", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "bd9f369a13c340e72c29924756cd8736d551219a5ab0a37c62e83e6eebfbe35a", "generation_seconds_batch": 23.611274659982882, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.105729103088379, "target_z": 1.0937499027280755, "ce": 6.348263263702393, "genuineness": 4.519899368286133, "identity": 0.5988713502883911, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4b5c8df10a9ce7d470891db09c58b0de15ade72b7302ff37ea7c19bc6d28daf6.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "4b73004eebabc34839c169a38c78d1fc582588e12e8fed70edfb627518c65985", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c966b3a0b41eab491f8b4b93e3c9c55b58109a085e37dd81fba9ca8ebb7763bc", "generation_seconds_batch": 22.20864488906227, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.551011085510254, "target_z": 0.6914062459470032, "ce": 6.165581703186035, "genuineness": 3.8671910762786865, "identity": 0.29792630672454834, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4b73004eebabc34839c169a38c78d1fc582588e12e8fed70edfb627518c65985.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "4ba1b136bea16875b61b6970278281310d97c5fa409a0f4283399041fc7f4a53", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "cf9be32104377a49c7c9af14567c11973c6931fa5de03316bf9947fdc7c7f71f", "generation_seconds_batch": 22.83592540200334, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4934933185577393, "target_z": 1.3749999567680335, "ce": 6.0146708488464355, "genuineness": 4.632867813110352, "identity": 0.6564561128616333, "hypothesis": "Die Tür ist noch offen. Ich hab deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.05263157894736842, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4ba1b136bea16875b61b6970278281310d97c5fa409a0f4283399041fc7f4a53.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "4ce21c112cec018984327c99c0fe39ae2d21f8eeafd3f6f63d58755100702452", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "05631db3196133f2507995c53631ecacd5457f548ebebc72b59b36fbab0b3832", "generation_seconds_batch": 24.453602167079225, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.987245798110962, "target_z": 1.0078125351259728, "ce": 6.2767791748046875, "genuineness": 4.7081804275512695, "identity": 0.3148338794708252, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4ce21c112cec018984327c99c0fe39ae2d21f8eeafd3f6f63d58755100702452.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "4d4f7e50d81f1e6d18d0e3319eebf3e9c270e6478cda6ddf660af9ece4a50b28", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d0faa4f556c230b45bc9219c6533413670f8161b0a2a5610538bef9fdd5bd92c", "generation_seconds_batch": 24.952836701995693, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.793363571166992, "target_z": 0.8671874216420608, "ce": 6.395919322967529, "genuineness": 4.457139015197754, "identity": 0.20940649509429932, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4d4f7e50d81f1e6d18d0e3319eebf3e9c270e6478cda6ddf660af9ece4a50b28.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "4d90a22d60e7b5fdcccc9d779514de7af0cf0fdc42b3fc3646c71a3be1f16c09", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "12f900cd1971bee6c2701241a143ef11f143eb4da45ac87e4785fc5b7feb4119", "generation_seconds_batch": 17.110385830979794, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4900453090667725, "target_z": -0.0781250054039958, "ce": 6.2767791748046875, "genuineness": 4.432034492492676, "identity": 0.7027024030685425, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4d90a22d60e7b5fdcccc9d779514de7af0cf0fdc42b3fc3646c71a3be1f16c09.mp3" }, { "id": "heldout-de-1", "language": "de", "sentences": [ "Die Schlüssel liegen auf dem Küchentisch.", "Ich suchte in der Schublade... dann neben dem Fenster.", "Nimm sie mit, wenn du gehst." ], "text": "Die Schlüssel liegen auf dem Küchentisch. Ich suchte in der Schublade... dann neben dem Fenster. Nimm sie mit, wenn du gehst.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-1", "key": "4dbdce2d8f2d7c0d7e1ca32670ae6da9d85ad3d3a5be5e1f2dba4d272c9c64ab", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Natural clear speech.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Natural clear speech.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b8dac2511f129c4178c51a31d0a4db3d3998ca13114b0907562cbbebab77fd93", "generation_seconds_batch": 15.825846425956115, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.139766693115234, "target_z": 1.8437498162641426, "ce": 5.6334228515625, "genuineness": 4.394378662109375, "identity": 0.5474291443824768, "hypothesis": "Die Schlüssel liegen auf dem Küchentisch. Ich suchte in der Schublade, dann neben dem Fenster. Nimm sie mit, wenn du gehst.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4dbdce2d8f2d7c0d7e1ca32670ae6da9d85ad3d3a5be5e1f2dba4d272c9c64ab.mp3" }, { "id": "heldout-nl-1", "language": "nl", "sentences": [ "De sleutels liggen op de keukentafel.", "Ik keek in de la... en daarna naast het raam.", "Neem ze mee als je weggaat." ], "text": "De sleutels liggen op de keukentafel. Ik keek in de la... en daarna naast het raam. Neem ze mee als je weggaat.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-1", "key": "4e0f0322640a94237f6985fa9780ccfae37037be06a23f00f8ca39dabdd66171", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Neem ze mee als je weggaat.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Natural clear speech.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Natural clear speech.) [3.2 seconds duration] Neem ze mee als je weggaat.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "076552b525d203528c81f38b24b68195fb7707781646c87c714e607d627147a1", "generation_seconds_batch": 15.059516415931284, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5365781784057617, "target_z": 1.4062499243440587, "ce": 5.863759994506836, "genuineness": 4.118232727050781, "identity": 0.5895484089851379, "hypothesis": "De sleutels liggen op de keukentafel. Ik keek in de la en daarna, naast het raam, neem ze mee als je weggaat.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4e0f0322640a94237f6985fa9780ccfae37037be06a23f00f8ca39dabdd66171.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "4e1217b607b9707a12f15c7677a6078581c1e05bd43b0a6bef0b4107e4be1a08", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1e2ced4d1e8af7c7083eff746c3e25f8cbb61edcd08a5b2f66f492063070b46f", "generation_seconds_batch": 24.118244329933077, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3836795091629028, "target_z": -0.15527340068527853, "ce": 6.284721851348877, "genuineness": 4.457139015197754, "identity": 0.5775860548019409, "hypothesis": "ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4e1217b607b9707a12f15c7677a6078581c1e05bd43b0a6bef0b4107e4be1a08.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "4e13d6fa4e32fba1d2fedde2f8d4893bc385528c42af709fdc7cef517d86b032", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "c542bb898bd4982345df8c6cbe2649ee6a4c30f4c088b494470935353c8bf804", "generation_seconds_batch": 27.643583735916764, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.2259368896484375, "target_z": 1.9062500972719245, "ce": 5.434855937957764, "genuineness": 4.494794845581055, "identity": 0.6479324102401733, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle comprend.", "wer": 0.05263157894736842, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4e13d6fa4e32fba1d2fedde2f8d4893bc385528c42af709fdc7cef517d86b032.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "4e5ddd6c32b2d8564e40329c927884dbd8b01e77b9d36083ba305519f3cb3f2b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8cc9733d7c337bb978cd3b6722012edebb215c8d38ada89dd4fd9f8ec5d9a414", "generation_seconds_batch": 23.07523575297091, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9135711193084717, "target_z": 1.679687443258044, "ce": 5.446769714355469, "genuineness": 4.20609712600708, "identity": 0.19189783930778503, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4e5ddd6c32b2d8564e40329c927884dbd8b01e77b9d36083ba305519f3cb3f2b.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "4e79980d9d15b22bd8333a6d63c1ec24a23338efd5ea04348e6fb3d40ab7f3eb", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "edad8bc5d1dbad235f04ae78a988f85405d5ec6c2a6a8bf7f1a26eb58145b1d2", "generation_seconds_batch": 17.018014376983047, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7950878143310547, "target_z": 1.5937500756559413, "ce": 5.903473377227783, "genuineness": 4.7332844734191895, "identity": 0.5329636335372925, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4e79980d9d15b22bd8333a6d63c1ec24a23338efd5ea04348e6fb3d40ab7f3eb.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "4ebb32535a9d3f8c557e3c6bb78a4a3c7b4ce29af93142ec99b064610fbd9f53", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ca4136b2650d01c9fd48545027b453f9d1e6f6c5fe2aa886e9a5eda7df7b66f4", "generation_seconds_batch": 17.07711539103184, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.549073219299316, "target_z": 2.14062476762818, "ce": 5.657250881195068, "genuineness": 4.607763767242432, "identity": 0.8641477227210999, "hypothesis": "La porte est encore ouverte, j'ai posé ton manteau sur la chaise, dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/4ebb32535a9d3f8c557e3c6bb78a4a3c7b4ce29af93142ec99b064610fbd9f53.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "505ed09f2a1cc70965aec6c6bca4d298ab7ddaa2237dd6c073ed544794992db8", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b134b1fc60e9d2eb6764a399350c262c968d0ebb6c0813dac546844e261a8b8e", "generation_seconds_batch": 14.67994263197761, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5654438734054565, "target_z": -0.023437518913985333, "ce": 6.268836498260498, "genuineness": 4.469690799713135, "identity": 0.51082444190979, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/505ed09f2a1cc70965aec6c6bca4d298ab7ddaa2237dd6c073ed544794992db8.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "50d476a5860466eac8d0e7b320eb2398af24516311a0236b110958eb0cc5ffd9", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d4f906878041d9afb9cbd6fc9130ee787cf88af7a31b54c7cda8e1e22b7919f3", "generation_seconds_batch": 22.618776715011336, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.524083137512207, "target_z": 0.6718750810599371, "ce": 6.046441555023193, "genuineness": 4.15588903427124, "identity": 0.47109174728393555, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/50d476a5860466eac8d0e7b320eb2398af24516311a0236b110958eb0cc5ffd9.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "5184af93cd1a738ce253d9c5552656323cc84313e2454d70b18d1eda5f509af7", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8829a5e555fcb952568e7785d2eced641a58eafb69cb48d072dac81f40c9972e", "generation_seconds_batch": 24.786545202950947, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5365781784057617, "target_z": 1.4062499243440587, "ce": 6.1020402908325195, "genuineness": 4.570107460021973, "identity": 0.5806903839111328, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/5184af93cd1a738ce253d9c5552656323cc84313e2454d70b18d1eda5f509af7.mp3" }, { "id": "heldout-de-3", "language": "de", "sentences": [ "Jemand hat den kleinen blauen Stuhl umgestellt.", "Er stand hier... neben dem hohen Bücherregal.", "Wir stellen ihn nach dem Essen zurück." ], "text": "Jemand hat den kleinen blauen Stuhl umgestellt. Er stand hier... neben dem hohen Bücherregal. Wir stellen ihn nach dem Essen zurück.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-3", "key": "52020ab92d7175b9c006b021752434fc007c050d342bea3e06f65f8f8a6a9cb0", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Natural clear speech.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Natural clear speech.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8a279223fbadd317d1588ca6425d5fd3b62214df1d5a4595409caf1c203cc255", "generation_seconds_batch": 16.33444074296858, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.95493221282959, "target_z": 0.9843751026759204, "ce": 5.506340026855469, "genuineness": 4.243753433227539, "identity": 0.36510324478149414, "hypothesis": "Jemand hat den kleinen blauen Stuhl umgestellt. Er stand hier neben dem hohen Bücherregal. Wir stellen ihn nach dem Essen zurück.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/52020ab92d7175b9c006b021752434fc007c050d342bea3e06f65f8f8a6a9cb0.mp3" }, { "id": "heldout-de-3", "language": "de", "sentences": [ "Jemand hat den kleinen blauen Stuhl umgestellt.", "Er stand hier... neben dem hohen Bücherregal.", "Wir stellen ihn nach dem Essen zurück." ], "text": "Jemand hat den kleinen blauen Stuhl umgestellt. Er stand hier... neben dem hohen Bücherregal. Wir stellen ihn nach dem Essen zurück.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-3", "key": "52c0a7a3c21c1fb3aaf6c404c4b96ea377bb01ceb7db3ab94d7a9b068ddf2fc9", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Natural clear speech.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Natural clear speech.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "bc4dc8fbd08148b5e289c93f9f2d5323bf0f29bc18d6168d7a079e037133389c", "generation_seconds_batch": 16.33444074296858, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.202670097351074, "target_z": 1.1640623730060984, "ce": 5.903473377227783, "genuineness": 4.482243061065674, "identity": 0.43839240074157715, "hypothesis": "Jemand hat den kleinen blauen Stuhl umgestellt. Er stand hier, neben dem hohen Bücherregal. Wir stellen ihn nach dem Essen zurück.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/52c0a7a3c21c1fb3aaf6c404c4b96ea377bb01ceb7db3ab94d7a9b068ddf2fc9.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "52cde0349ad8a4a741efde5eaa9c28a7063a2589c179f8573cfc7744cb4295ee", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3338938319718a0c0bf5c0e9e0b743c83b6bc5b520c6f8c1c0a5b51cf6084794", "generation_seconds_batch": 24.557565840077586, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5581207275390625, "target_z": 1.421874994596004, "ce": 5.8875885009765625, "genuineness": 4.394378662109375, "identity": 0.5183687806129456, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hinein.", "wer": 0.0625, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/52cde0349ad8a4a741efde5eaa9c28a7063a2589c179f8573cfc7744cb4295ee.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "52f8ca98edc35a0b5391945072b08aa11e10350e480fb0822c09da232e805557", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f8f6d34881f7ba78a5071cbb6fce783821e0a8a001fc0e45a99de5763ed51c3a", "generation_seconds_batch": 19.000402122968808, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.960317611694336, "target_z": 0.9882811973110408, "ce": 6.348263263702393, "genuineness": 4.7458367347717285, "identity": 0.777478814125061, "hypothesis": "Te he encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.058823529411764705, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/52f8ca98edc35a0b5391945072b08aa11e10350e480fb0822c09da232e805557.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "532c6332e1067d0c1ec88642483e2791dd6bf5a07b2186f9b14529b6eb7d1b4e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d0279a33704fc60e7c62d910ea16d54e39674175a340510278f3bfe558626bb7", "generation_seconds_batch": 23.278622932964936, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1703567504882812, "target_z": 1.140625113483912, "ce": 6.09409761428833, "genuineness": 4.344170093536377, "identity": 0.6428486108779907, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/532c6332e1067d0c1ec88642483e2791dd6bf5a07b2186f9b14529b6eb7d1b4e.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "538b5e9b10a4e52ac0b71777f372e38d23892e1d1673c91fc8ceac9c4927cabc", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "317ab3a414caf520d8eb52ab372e4c4285c06bbd9360f848db4376ada5160c9b", "generation_seconds_batch": 31.204979359987192, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3032729625701904, "target_z": 0.5117188026889592, "ce": 6.4515180587768555, "genuineness": 4.632867813110352, "identity": 0.3483179807662964, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/538b5e9b10a4e52ac0b71777f372e38d23892e1d1673c91fc8ceac9c4927cabc.mp3" }, { "id": "heldout-de-1", "language": "de", "sentences": [ "Die Schlüssel liegen auf dem Küchentisch.", "Ich suchte in der Schublade... dann neben dem Fenster.", "Nimm sie mit, wenn du gehst." ], "text": "Die Schlüssel liegen auf dem Küchentisch. Ich suchte in der Schublade... dann neben dem Fenster. Nimm sie mit, wenn du gehst.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-1", "key": "5432c641cee9c761fbc1b3215044883c8423627b6d24cc080c77c6a671231b38", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Natural clear speech.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Natural clear speech.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "63d5e54168db2203a3e8804cd20cd4defeb334c6aec789734415fce382b3abbf", "generation_seconds_batch": 23.010773499961942, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7950878143310547, "target_z": 1.5937500756559413, "ce": 5.982900142669678, "genuineness": 4.432034492492676, "identity": 0.4843348264694214, "hypothesis": "Die Schlüssel liegen auf dem Küchentisch. Ich suchte in der Schublade, dann neben dem Fenster. Nimm sie mit, wenn du gehst.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/5432c641cee9c761fbc1b3215044883c8423627b6d24cc080c77c6a671231b38.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "545b5ea5fcadaaebd2131885084e576b4e8c94d7c3e36ee4969a46a8040f9847", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "927ae323a2c24ee5c77ae1f4f92a43fcdba3e5aa264d024b576a038b45c3ce43", "generation_seconds_batch": 22.83836459298618, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.51503586769104, "target_z": 1.3906250270199791, "ce": 5.530168056488037, "genuineness": 4.168440818786621, "identity": 0.3454720377922058, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/545b5ea5fcadaaebd2131885084e576b4e8c94d7c3e36ee4969a46a8040f9847.mp3" }, { "id": "heldout-de-4", "language": "de", "sentences": [ "Ich schrieb die Nummer auf den Umschlag.", "Lies sie genau... und prüfe die letzten beiden Ziffern.", "Ruf an, wenn du Zeit hast." ], "text": "Ich schrieb die Nummer auf den Umschlag. Lies sie genau... und prüfe die letzten beiden Ziffern. Ruf an, wenn du Zeit hast.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-4", "key": "546203f143d2dbd908be3097fd76d41d76e93afd1487fa15c64a3d119dff5d49", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d8df71efaeba0bfcdd05b9ee818b74b2d29cf8b2d801cab367893e47e6eca1a3", "generation_seconds_batch": 17.61561208299827, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1507517099380493, "target_z": -0.3242187378410094, "ce": 6.24500846862793, "genuineness": 4.086852550506592, "identity": 0.5894517302513123, "hypothesis": "Ich schrieb die Nummer auf den Umschlag, ließ sie genau und prüfe die letzten beiden Ziffern. Ruf an, wenn du Zeit hast.", "wer": 0.045454545454545456, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/546203f143d2dbd908be3097fd76d41d76e93afd1487fa15c64a3d119dff5d49.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "549ae57dd5b3e955e94f007a94996de891e5b558e19e30a7d591ebfd93147228", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7b18cc88032fb402e905de0b450959c2cb35cfbcf2a415e586693379c3c099f9", "generation_seconds_batch": 17.416993562947027, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5581207275390625, "target_z": 1.421874994596004, "ce": 5.5539960861206055, "genuineness": 4.124508857727051, "identity": 0.26081663370132446, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/549ae57dd5b3e955e94f007a94996de891e5b558e19e30a7d591ebfd93147228.mp3" }, { "id": "heldout-de-2", "language": "de", "sentences": [ "Die Straße führt hinter der Brücke nach links.", "Geh weiter... bis zu dem kleinen Holztor.", "Ich warte auf der anderen Seite." ], "text": "Die Straße führt hinter der Brücke nach links. Geh weiter... bis zu dem kleinen Holztor. Ich warte auf der anderen Seite.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-2", "key": "556d0089965bb9ba3ce35a256c7c9dc5c382caa0198f45955e87a32037ca34aa", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Natural clear speech.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Natural clear speech.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f22a9423e91b21546694822e66018b44fb9494fcbbef2c464ea2b622a40fa8ff", "generation_seconds_batch": 16.11652067489922, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.342696189880371, "target_z": 1.2656249837880125, "ce": 5.470598220825195, "genuineness": 4.256305694580078, "identity": 0.4405738115310669, "hypothesis": "Die Straße führt hinter der Brücke nach links. Geh weiter bis zu dem kleinen Holztor. Ich warte auf der anderen Seite.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/556d0089965bb9ba3ce35a256c7c9dc5c382caa0198f45955e87a32037ca34aa.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "56411971a3daeded82bd5097e546ca13b408aef3f743ffaa882009227846f260", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8d66f9cc17a6516a4e7941059b521937e12ece80292dda8db04e25b7fca1aefa", "generation_seconds_batch": 24.26231643895153, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8166301250457764, "target_z": 1.6093749729800209, "ce": 5.510311126708984, "genuineness": 4.507347106933594, "identity": 0.4826875329017639, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/56411971a3daeded82bd5097e546ca13b408aef3f743ffaa882009227846f260.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "56a32e3686213d4502d50903734c91e9f748c0cc8119d2d39b654b60bd490a85", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "9abeece9fea45736b26a0ec23f4f91b2fee522722ac2329455a65558c5e1e18c", "generation_seconds_batch": 18.49265929707326, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.633519411087036, "target_z": 1.4765625675499476, "ce": 5.613565921783447, "genuineness": 4.055472373962402, "identity": 0.5698741674423218, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás ligado.", "wer": 0.06666666666666667, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/56a32e3686213d4502d50903734c91e9f748c0cc8119d2d39b654b60bd490a85.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "56a9f5b5a9de01b20dd9864a13bc2326d0dab3939bdc297682d8808889547324", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b93cda9b8303a91d3040b68f4d471db0fe444ad95228f0cae3e1ac74bcdd21db", "generation_seconds_batch": 23.261335554067045, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2611567974090576, "target_z": -0.24414059865552043, "ce": 6.284721851348877, "genuineness": 4.494794845581055, "identity": 0.6754093170166016, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg, maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/56a9f5b5a9de01b20dd9864a13bc2326d0dab3939bdc297682d8808889547324.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "570040aa06ea06fe6bdcc8acc8522e40bff28aea4f0f6e0903cb0fbfad924910", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5c6b785f8a6cd69e2c579864a67b9ecfcfe7b639671186402a1625a1ea8b1d0b", "generation_seconds_batch": 25.116734603070654, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8364486694335938, "target_z": 0.8984375621459518, "ce": 6.205295085906982, "genuineness": 4.7332844734191895, "identity": 0.8043028116226196, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/570040aa06ea06fe6bdcc8acc8522e40bff28aea4f0f6e0903cb0fbfad924910.mp3" }, { "id": "heldout-fr-4", "language": "fr", "sentences": [ "J’ai écrit le numéro sur l’enveloppe.", "Lis-le bien... et vérifie les deux derniers chiffres.", "Appelle-moi quand tu auras un moment." ], "text": "J’ai écrit le numéro sur l’enveloppe. Lis-le bien... et vérifie les deux derniers chiffres. Appelle-moi quand tu auras un moment.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-4", "key": "570c24da4752c19c73405cdbd3ea7f27889d5ea1da2fa140bcd7d7155cbab106", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Natural clear speech.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Natural clear speech.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3d767ed6da269741c8533d2210fa839f4d738eef6aa33b57bc91073ede03c9aa", "generation_seconds_batch": 17.38643721397966, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7950878143310547, "target_z": 1.5937500756559413, "ce": 5.839932441711426, "genuineness": 4.532451152801514, "identity": 0.7969386577606201, "hypothesis": "J'ai écrit le numéro sur l'enveloppe, lis-le bien et vérifie les deux derniers chiffres. Appelle-moi quand il a un moment.", "wer": 0.09090909090909091, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/570c24da4752c19c73405cdbd3ea7f27889d5ea1da2fa140bcd7d7155cbab106.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "57403f833f94422808e435874f98326b92dac82643cbee83dd5e4ecb4a30c5f2", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "e7d4d18535adbb95edd83ad69e391ffbe37214b2d45f2abd6ee6bc233f95ff44", "generation_seconds_batch": 22.47322924400214, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.224212646484375, "target_z": 1.179687443258044, "ce": 6.117925643920898, "genuineness": 4.7081804275512695, "identity": 0.6387518048286438, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/57403f833f94422808e435874f98326b92dac82643cbee83dd5e4ecb4a30c5f2.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "57425b67a22c452daae80c071134704a663d5aec85cc934ae6ffa922f34b95db", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "d84e11b7a82d32c469202213430d954cd3f6d2c221919bbe1f6b477de013ae54", "generation_seconds_batch": 17.058061757939868, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.428865909576416, "target_z": 1.3281249189400628, "ce": 5.982900142669678, "genuineness": 4.519899368286133, "identity": 0.6884416937828064, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/57425b67a22c452daae80c071134704a663d5aec85cc934ae6ffa922f34b95db.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "57901f301e10e0e9acdcae0b80e677946327197e88fa2c8d948290c792291619", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "81c4aa236e7ced190d62c564c7bfa076fe306aaeb6db9a1f13d4af20d64c4c23", "generation_seconds_batch": 23.523478726041503, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.171325445175171, "target_z": 0.41601567971545755, "ce": 6.221180438995361, "genuineness": 4.375550270080566, "identity": 0.4322184920310974, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/57901f301e10e0e9acdcae0b80e677946327197e88fa2c8d948290c792291619.mp3" }, { "id": "heldout-de-1", "language": "de", "sentences": [ "Die Schlüssel liegen auf dem Küchentisch.", "Ich suchte in der Schublade... dann neben dem Fenster.", "Nimm sie mit, wenn du gehst." ], "text": "Die Schlüssel liegen auf dem Küchentisch. Ich suchte in der Schublade... dann neben dem Fenster. Nimm sie mit, wenn du gehst.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-1", "key": "57b016933e8918aa870423cbc2939dd4f3108ec7c4e1829ac18400cf578e4d32", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Natural clear speech.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Natural clear speech.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0d23f12428d74f1147a11be2816955c2701a06c6c4ba77376399601a7b160b1d", "generation_seconds_batch": 15.825846425956115, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.741231679916382, "target_z": 1.5546875729539433, "ce": 5.681078910827637, "genuineness": 4.193545341491699, "identity": 0.5090473890304565, "hypothesis": "Die Schlüssel liegen auf dem Küchentisch. Ich suchte in der Schublade, dann neben dem Fenster. Nimm sie mit, wenn du gehst.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/57b016933e8918aa870423cbc2939dd4f3108ec7c4e1829ac18400cf578e4d32.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "57de4552ded5037653dc214216ec783a99166b07a7227f0eac961e7ac0547839", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "89597e9ead02f4552e4c3579e4750fb8dae60585a558312a85dfacc3d8260f86", "generation_seconds_batch": 23.293156858999282, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8166301250457764, "target_z": 1.6093749729800209, "ce": 6.022613525390625, "genuineness": 4.532451152801514, "identity": 0.6743124723434448, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/57de4552ded5037653dc214216ec783a99166b07a7227f0eac961e7ac0547839.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "57df67064ca5eb8daad08b7f09eeba379de7e4542adc39f4c2dc9270e2345a11", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "6e246a06190a567fc41abd5a9650659ef72fae09337e901d4d1a4a8bb0b2cd42", "generation_seconds_batch": 24.345523774973117, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6766042709350586, "target_z": 1.5078125351259728, "ce": 5.202532768249512, "genuineness": 4.017816066741943, "identity": 0.49097201228141785, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás liquido.", "wer": 0.06666666666666667, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/57df67064ca5eb8daad08b7f09eeba379de7e4542adc39f4c2dc9270e2345a11.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "57e9264fe8be0f777dcb1b362fd8388c004c80d81da247e1bffd94df279634a4", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ff85a052a0d8e07047a2c8428f9fc95205442ec75a449a064da46b4a3a40017f", "generation_seconds_batch": 25.43855618208181, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3086585998535156, "target_z": 0.5156250702519455, "ce": 6.356205940246582, "genuineness": 4.519899368286133, "identity": 0.36118245124816895, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/57e9264fe8be0f777dcb1b362fd8388c004c80d81da247e1bffd94df279634a4.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "584100ce519f3430c08edfdc36e4450bf8554108bbad1ce9a65233391c688057", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "37c55509c2ef68c5792d228d08ce2ce3e169c62d4f4d9e4503aa2ec6008b0549", "generation_seconds_batch": 17.252815083018504, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.120162010192871, "target_z": 0.3789062243310199, "ce": 6.173524379730225, "genuineness": 4.432034492492676, "identity": 0.8295215368270874, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/584100ce519f3430c08edfdc36e4450bf8554108bbad1ce9a65233391c688057.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "58610e11b17057ccda639b808212fd742c262a25ec24a3a96a5cac493bdc551a", "bucket": "vn_S_ASMR_reg", "phase": "wording", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "b7e1f12fc4debe44f7be4410a7a73614ef9aad3e94d2319b6ce8d1aa20a8107a", "generation_seconds_batch": 17.731394647969864, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.939743995666504, "target_z": 0.2480468662185068, "ce": 6.197352409362793, "genuineness": 4.432034492492676, "identity": 0.8192926645278931, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/58610e11b17057ccda639b808212fd742c262a25ec24a3a96a5cac493bdc551a.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "58af6c0b9282da5bd4f9ba373f8ceaa79d28453ad3c01afa19e71ef4ec251efb", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 130, "seconds_generated": 10.4, "mp3_sha256": "1e47d69290a24e3856ea7230cd14b308964a858302981f6ffd7b998aa16fc80a", "generation_seconds_batch": 24.345523774973117, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.859714984893799, "target_z": 1.640624940556046, "ce": 5.283945083618164, "genuineness": 4.2751336097717285, "identity": 0.32193177938461304, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás.", "wer": 0.06666666666666667, "duration_error_s": 0.20000000000000107, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/58af6c0b9282da5bd4f9ba373f8ceaa79d28453ad3c01afa19e71ef4ec251efb.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "58b17cb2db6f43b88c4217fbb99b119f31b0f0757a25dd25b8c98e25aadbf73d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3d831ea164dcc3c1c90a191d9b046b00112c4ddb4fdf95781d215c0fca1a0533", "generation_seconds_batch": 24.082210180000402, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.203639030456543, "target_z": 0.43945311216551, "ce": 6.181467056274414, "genuineness": 4.482243061065674, "identity": 0.5376467108726501, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/58b17cb2db6f43b88c4217fbb99b119f31b0f0757a25dd25b8c98e25aadbf73d.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "58ccb3dcf19b9067f34a5ef95a6420cfa6dd75928adaebe525cb45b450b3b213", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1087bf85cc87d7e34fcdcaeff7e9204319456207ffdaa765aa86beb13f7cd122", "generation_seconds_batch": 23.255606304039247, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.701808214187622, "target_z": 0.8007812189270241, "ce": 6.300607204437256, "genuineness": 4.457139015197754, "identity": 0.6147308349609375, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/58ccb3dcf19b9067f34a5ef95a6420cfa6dd75928adaebe525cb45b450b3b213.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "592f8db1eb95cae13baca496af70d15befa07ac1999bf5259e364f05b56ae4e9", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f4bda52c5afec8fccb6eb760d727bf3c7774c8a3abb1a1c7f5c8c9090d9854df", "generation_seconds_batch": 24.121383397025056, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.051873207092285, "target_z": 1.0546875729539433, "ce": 6.284721851348877, "genuineness": 4.519899368286133, "identity": 0.6434115767478943, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschneben.", "wer": 0.0625, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/592f8db1eb95cae13baca496af70d15befa07ac1999bf5259e364f05b56ae4e9.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "597010d7d17fec4c4e8b2b2646fe2e935de2bb2657d795b89b702a2c22bd4275", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d7a5828b90d1bf8c5c1c1e891e93e252279fcbd8d7db47fa4019f55ac1bf1bcf", "generation_seconds_batch": 25.812747161951847, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3625144958496094, "target_z": 0.5546874000260775, "ce": 6.252951145172119, "genuineness": 4.607763767242432, "identity": 0.48131585121154785, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/597010d7d17fec4c4e8b2b2646fe2e935de2bb2657d795b89b702a2c22bd4275.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "59786f3a5cbe7dba3501dc910f9f2103137e848fc1267968b9c27744090265d5", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "04de50c8119a5d45c3c295488cbd1761a118b0c0252d624c7f82b69f2be77da0", "generation_seconds_batch": 25.363981387927197, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7089178562164307, "target_z": 1.5312499675760252, "ce": 5.546053409576416, "genuineness": 4.111956596374512, "identity": 0.4242818355560303, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/59786f3a5cbe7dba3501dc910f9f2103137e848fc1267968b9c27744090265d5.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "5989b693915e8fea87aaa31b312492455ff4b0f2a6cb0c8909279de7229a15ed", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "91da0e1269bdc80d4403204dc84eaea715833835deee6d1f92f8dbdba307595e", "generation_seconds_batch": 23.144209127989598, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.245755195617676, "target_z": 1.1953125135099896, "ce": 6.09409761428833, "genuineness": 4.582659721374512, "identity": 0.6964966058731079, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/5989b693915e8fea87aaa31b312492455ff4b0f2a6cb0c8909279de7229a15ed.mp3" }, { "id": "heldout-nl-3", "language": "nl", "sentences": [ "Iemand heeft het kleine blauwe stoeltje verplaatst.", "Het stond hier... naast de hoge boekenkast.", "We zetten het na de lunch terug." ], "text": "Iemand heeft het kleine blauwe stoeltje verplaatst. Het stond hier... naast de hoge boekenkast. We zetten het na de lunch terug.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-3", "key": "59e1ef2dce411461dc37bf354811637e89d18e53de26f69ccf258589460eaa5d", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We zetten het na de lunch terug.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Natural clear speech.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Natural clear speech.) [3.2 seconds duration] We zetten het na de lunch terug.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "dfd67d99e0bba370a9293e5cefc71d4aebbf13da1a6766c198ebe87cf8fea7a6", "generation_seconds_batch": 23.683933655032888, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0626444816589355, "target_z": 1.0625001080799161, "ce": 6.054384231567383, "genuineness": 4.582659721374512, "identity": 0.5873544216156006, "hypothesis": "Iemand heeft het kleine blauwe stoeltje verplaatst. Het stond hier, naast de hoge boekenkast. We zetten het naar de lunch terug.", "wer": 0.047619047619047616, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/59e1ef2dce411461dc37bf354811637e89d18e53de26f69ccf258589460eaa5d.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "5a07f695916ce8bf3b392f4636a8deae32523e505cbc4b09a6a992ba2609ffa1", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7ea666b51c4dd7609ba6794e410c7bfa0831f0825573f9879aea2076a47a47a3", "generation_seconds_batch": 25.344684287905693, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.868762254714966, "target_z": 0.9218749945960042, "ce": 6.109982967376709, "genuineness": 4.532451152801514, "identity": 0.7295715808868408, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/5a07f695916ce8bf3b392f4636a8deae32523e505cbc4b09a6a992ba2609ffa1.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "5ad6a8125a7cddc46bb54ddd80e31583977a6ad70a6219ecc5ab4515f1dcdc6e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "cb97a6c84a7ae6f5b2511868a46733198e43cdab5c01d9d59b640d6306e88ce8", "generation_seconds_batch": 25.484242056962103, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8633766174316406, "target_z": 0.9179687270330178, "ce": 6.149696350097656, "genuineness": 4.3881025314331055, "identity": 0.7016904354095459, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/5ad6a8125a7cddc46bb54ddd80e31583977a6ad70a6219ecc5ab4515f1dcdc6e.mp3" }, { "id": "heldout-nl-3", "language": "nl", "sentences": [ "Iemand heeft het kleine blauwe stoeltje verplaatst.", "Het stond hier... naast de hoge boekenkast.", "We zetten het na de lunch terug." ], "text": "Iemand heeft het kleine blauwe stoeltje verplaatst. Het stond hier... naast de hoge boekenkast. We zetten het na de lunch terug.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-3", "key": "5aefcae9ccc4bd29230ed598b6a56d9228a48f530835cacafa051fe428a3830e", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We zetten het na de lunch terug.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Natural clear speech.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Natural clear speech.) [3.2 seconds duration] We zetten het na de lunch terug.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f43e26cc47dbbc9af52dfd98a8600cbdf425c8993cd1cb3f12fc89744ebdb50c", "generation_seconds_batch": 14.943401627941057, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.610252857208252, "target_z": 0.7343750162119874, "ce": 5.935244083404541, "genuineness": 4.457139015197754, "identity": 0.5793063044548035, "hypothesis": "Iemand heeft het kleine blauwe stoeltje verplaatst. Het stond hier naast de hoge boekenkast. We zetten het naar de lunch terug.", "wer": 0.047619047619047616, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/5aefcae9ccc4bd29230ed598b6a56d9228a48f530835cacafa051fe428a3830e.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "5b5f2857f8bfdef92e13b06ead793e6665c1ce1a4eb5c8d75cd88b1f279cdd8b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d333a52b60296855429e7fc6335a7b5aa606c0f8dfc10c89b16f0e28e1cc9a3c", "generation_seconds_batch": 25.43855618208181, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1363189220428467, "target_z": 0.39062502701997903, "ce": 6.23706579208374, "genuineness": 4.174716949462891, "identity": 0.4046173393726349, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/5b5f2857f8bfdef92e13b06ead793e6665c1ce1a4eb5c8d75cd88b1f279cdd8b.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "5c807bdd6fbcfab3517df0bb15a59b7f4bf93a58472b7859827af13f8f8b331b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "e6667c746672e55e522fac7e20ae6a19b71a2bef94750cfdeb34d954868c0bcd", "generation_seconds_batch": 24.453602167079225, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1642156839370728, "target_z": -0.3144531553974764, "ce": 6.086154937744141, "genuineness": 4.557555198669434, "identity": 0.7201784253120422, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/5c807bdd6fbcfab3517df0bb15a59b7f4bf93a58472b7859827af13f8f8b331b.mp3" }, { "id": "heldout-de-2", "language": "de", "sentences": [ "Die Straße führt hinter der Brücke nach links.", "Geh weiter... bis zu dem kleinen Holztor.", "Ich warte auf der anderen Seite." ], "text": "Die Straße führt hinter der Brücke nach links. Geh weiter... bis zu dem kleinen Holztor. Ich warte auf der anderen Seite.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-2", "key": "5c814d6c57d97c60c9eddb92a573c6f4b99892c2982639094b030b5db4a2cb7a", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Natural clear speech.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Natural clear speech.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "bcf3596c9dc199ef323e6df1c7bc42ba53d7d81c7212155bf18df63b6af52168", "generation_seconds_batch": 16.11652067489922, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7089178562164307, "target_z": 1.5312499675760252, "ce": 5.792276382446289, "genuineness": 4.482243061065674, "identity": 0.3705245852470398, "hypothesis": "Die Straße führt hinter der Brücke nach links. Geh weiter bis zu dem kleinen Holztor. Ich warte auf der anderen Seite.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/5c814d6c57d97c60c9eddb92a573c6f4b99892c2982639094b030b5db4a2cb7a.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "5cb5efa1eec1c28c3413f980d0fc741f106bb007174c3cefec918ff8939c984c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "a0a9b78f1cbb1891f9e724d628513e618c04f18f53303b008d92e36ed34b4495", "generation_seconds_batch": 22.712467964971438, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.924342393875122, "target_z": 1.6874999783840168, "ce": 5.756534099578857, "genuineness": 4.319066047668457, "identity": 0.6650073528289795, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/5cb5efa1eec1c28c3413f980d0fc741f106bb007174c3cefec918ff8939c984c.mp3" }, { "id": "heldout-de-3", "language": "de", "sentences": [ "Jemand hat den kleinen blauen Stuhl umgestellt.", "Er stand hier... neben dem hohen Bücherregal.", "Wir stellen ihn nach dem Essen zurück." ], "text": "Jemand hat den kleinen blauen Stuhl umgestellt. Er stand hier... neben dem hohen Bücherregal. Wir stellen ihn nach dem Essen zurück.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-3", "key": "5ce454384b5dd02661383e1f00f0f08ef14a9757185d2509db2e0eccdeda5acb", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Natural clear speech.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Natural clear speech.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "24332979afb8846fdbb554fd6ef87ab2ad519618d0e715ed2ac579faab89c637", "generation_seconds_batch": 22.946463749045506, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4934933185577393, "target_z": 1.3749999567680335, "ce": 5.95112943649292, "genuineness": 4.425758361816406, "identity": 0.6171585917472839, "hypothesis": "Jemand hat den kleinen blauen Stuhl umgestellt. Er stand hier neben dem hohen Bücherregal. Wir stellen ihn nach dem Essen.", "wer": 0.047619047619047616, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/5ce454384b5dd02661383e1f00f0f08ef14a9757185d2509db2e0eccdeda5acb.mp3" }, { "id": "heldout-es-0", "language": "es", "sentences": [ "Tu mensaje llegó después de que me fuera.", "Lo leí dos veces... y guardé el teléfono.", "Dime cuándo podemos hablar." ], "text": "Tu mensaje llegó después de que me fuera. Lo leí dos veces... y guardé el teléfono. Dime cuándo podemos hablar.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-0", "key": "5d52d37cdff1441a54f835b9c7e8163dd3ff4653a28eabdebe76c9d57f6753d6", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo podemos hablar.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Natural clear speech.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo podemos hablar.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c25df1129b51af8e3522db39b5e362ab28733c57182039262b4552b08371be4b", "generation_seconds_batch": 17.429731165990233, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1703567504882812, "target_z": 1.140625113483912, "ce": 5.158848285675049, "genuineness": 4.168440818786621, "identity": 0.39961594343185425, "hypothesis": "Tu mensaje llegó después de que me fuera, lo leí dos veces y guardé el teléfono. Dime cuándo podemos hablar.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/5d52d37cdff1441a54f835b9c7e8163dd3ff4653a28eabdebe76c9d57f6753d6.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "5dcf1009ab9a22dc479f9c00ed0ed9cb1025aad00024d20f1fd8ce8ca95b7043", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b316ad0db371938961d4e2337325b7077efa033a73c4be1b0053e1c9d0b375ce", "generation_seconds_batch": 23.4669335790677, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6981465816497803, "target_z": 1.5234374324500524, "ce": 5.831989288330078, "genuineness": 4.243753433227539, "identity": 0.5433740019798279, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/5dcf1009ab9a22dc479f9c00ed0ed9cb1025aad00024d20f1fd8ce8ca95b7043.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "5dfda04bfe15a8e01d99d9c8a35446919bc2c2eea5b6cb0ee99eb0a5844fdf7e", "bucket": "vn_S_ASMR_reg", "phase": "wording", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e19167a638266a0b42bdcb4c796c599670a72cb377ca2aa8a817e20173d0af9e", "generation_seconds_batch": 18.048736938042566, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.090541124343872, "target_z": 0.35742183919852777, "ce": 6.1894097328186035, "genuineness": 4.645420074462891, "identity": 0.6928682327270508, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/5dfda04bfe15a8e01d99d9c8a35446919bc2c2eea5b6cb0ee99eb0a5844fdf7e.mp3" }, { "id": "heldout-fr-2", "language": "fr", "sentences": [ "La route tourne à gauche après le pont.", "Continue... jusqu’à la petite barrière en bois.", "Je t’attendrai de l’autre côté." ], "text": "La route tourne à gauche après le pont. Continue... jusqu’à la petite barrière en bois. Je t’attendrai de l’autre côté.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-2", "key": "5e0af124ad7421917e897f2876d49b25e213a1bc87e01fe43f09b9c255c2a32f", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9282db3bf823ae4b66ddc0fad43e7d3d1ff53088b141e4de1befe965e3bca26f", "generation_seconds_batch": 16.59471099707298, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.209024667739868, "target_z": 0.4433593797284963, "ce": 6.070269584655762, "genuineness": 4.444586753845215, "identity": 0.6699347496032715, "hypothesis": "La route tourne à gauche après le pont. Continue jusqu'à la petite barrière en bois. Je t'attendrai de l'autre côté.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/5e0af124ad7421917e897f2876d49b25e213a1bc87e01fe43f09b9c255c2a32f.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "5e22bdfba52f2c2d9fa8d07f863579c1cc6aa41c716a8c982d87397c1d90d830", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "a16c3d5ec478bb8621ecd4348219744ebd20a9d8c0acfad624628a08df878065", "generation_seconds_batch": 23.13790539803449, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.247479438781738, "target_z": 1.92187516752387, "ce": 5.895530700683594, "genuineness": 4.645420074462891, "identity": 0.4644632339477539, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/5e22bdfba52f2c2d9fa8d07f863579c1cc6aa41c716a8c982d87397c1d90d830.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "5e56f14ff4acc271139fe8f6dad88a44a47a21862c0f1a7fb9d944a35e733ef8", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e3a599971d5ac5fe545407cf78450cbf43ffe8a4989bcb725967b86193416d93", "generation_seconds_batch": 23.38123934203759, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1282405853271484, "target_z": 0.38476571213943245, "ce": 6.1894097328186035, "genuineness": 4.645420074462891, "identity": 0.6383600234985352, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/5e56f14ff4acc271139fe8f6dad88a44a47a21862c0f1a7fb9d944a35e733ef8.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "5ea687637ad4ff155a09a3112a14ab4b2e032bc0c919a363d080745d9a25c0f5", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "de9890490ef2641333d8ec8e102ab37331bf1ea1da5e48191722dea5a1ade2d4", "generation_seconds_batch": 25.292252992046997, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3211536407470703, "target_z": 1.249999913536067, "ce": 5.903473377227783, "genuineness": 4.595211505889893, "identity": 0.5840665102005005, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/5ea687637ad4ff155a09a3112a14ab4b2e032bc0c919a363d080745d9a25c0f5.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "5ebaf7e332f1ee773865a03b47f58ed4b575fcd5facc495c0b88f3258e66b143", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "bf12ded53e813b973ae72dcefaa7be8f976c5ab3b2e1a83a0d15a83524a659e7", "generation_seconds_batch": 22.669165586004965, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.987245798110962, "target_z": 1.0078125351259728, "ce": 6.292664527893066, "genuineness": 4.419482707977295, "identity": 0.7024165391921997, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/5ebaf7e332f1ee773865a03b47f58ed4b575fcd5facc495c0b88f3258e66b143.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "5ee8727684871c63cf1f5fda7573aca5cd457458b6d88b96db45359416561756", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6e351d2db299105f34f3ca7a390da906bde7eb3b7806f3babd33b1c4ed70f3f2", "generation_seconds_batch": 25.092228548950516, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.576215147972107, "target_z": -0.01562498378801257, "ce": 6.197352409362793, "genuineness": 4.293961524963379, "identity": 0.6003754138946533, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/5ee8727684871c63cf1f5fda7573aca5cd457458b6d88b96db45359416561756.mp3" }, { "id": "heldout-fr-4", "language": "fr", "sentences": [ "J’ai écrit le numéro sur l’enveloppe.", "Lis-le bien... et vérifie les deux derniers chiffres.", "Appelle-moi quand tu auras un moment." ], "text": "J’ai écrit le numéro sur l’enveloppe. Lis-le bien... et vérifie les deux derniers chiffres. Appelle-moi quand tu auras un moment.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-4", "key": "5f9ef058e04eee5650a423562a885a6d78c2bcecc265d8add4c55fa6b8e3ccc0", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Natural clear speech.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Natural clear speech.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "cb746e61263b309255578b339a424276ff3a83129994585d6fda6fea9e8d79e1", "generation_seconds_batch": 15.1527314389823, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.096682071685791, "target_z": 1.8125000216159832, "ce": 5.677107334136963, "genuineness": 4.3881025314331055, "identity": 0.6526321172714233, "hypothesis": "J'ai écrit le numéro sur l'enveloppe, lime-le bien et vérifie les deux derniers vivres. Appelle-moi quand tu auras un moment.", "wer": 0.09090909090909091, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/5f9ef058e04eee5650a423562a885a6d78c2bcecc265d8add4c55fa6b8e3ccc0.mp3" }, { "id": "heldout-fr-0", "language": "fr", "sentences": [ "Ton message est arrivé après mon départ.", "Je l’ai relu... puis j’ai rangé mon téléphone.", "Dis-moi quand nous pourrons parler." ], "text": "Ton message est arrivé après mon départ. Je l’ai relu... puis j’ai rangé mon téléphone. Dis-moi quand nous pourrons parler.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-0", "key": "603e5f5518bfaad55eced896d64af2ac78a1823cca665b20f0b31190d10aad32", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Natural clear speech.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2111cb14a1f49d4ca1dddd12b2ebaf60262205ddc82b6f3bbc68f95c144305be", "generation_seconds_batch": 18.42447387694847, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.107453346252441, "target_z": 1.820312556741956, "ce": 5.736677646636963, "genuineness": 4.582659721374512, "identity": 0.7471789121627808, "hypothesis": "Ton message est arrivé après mon départ. Je l'ai relu, puis j'ai rangé mon téléphone. Dis-moi quand nous pourrons parler.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/603e5f5518bfaad55eced896d64af2ac78a1823cca665b20f0b31190d10aad32.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "60467223ab2f3c7727f32d0521aaa6338856175f3eeb5334f8b0ec5309f410ef", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "c556f6d612d9b66acd38057205662cbf79050cd978bd4a39129bc906d1d79f60", "generation_seconds_batch": 16.064213418052532, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.549073219299316, "target_z": 2.14062476762818, "ce": 5.708878040313721, "genuineness": 4.68307638168335, "identity": 0.5117858648300171, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana, vieja. Podemos mirar juntos lo que ha", "wer": 0.11764705882352941, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/60467223ab2f3c7727f32d0521aaa6338856175f3eeb5334f8b0ec5309f410ef.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "604d93e272ce9d2c2f1cdf4e0fec214506ecb76ca7736ef296bb373a4a9f6484", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6121887cf72bdc464397c8d1dd3441ea4a3d66b8cb41ce16e8fa42afd286299b", "generation_seconds_batch": 22.23857655900065, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5456254482269287, "target_z": 0.6874999783840168, "ce": 6.348263263702393, "genuineness": 4.331617832183838, "identity": 0.42340701818466187, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/604d93e272ce9d2c2f1cdf4e0fec214506ecb76ca7736ef296bb373a4a9f6484.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "60858eb98393dad6c71939ddd09e6ef5c6d4afdf565ebb80211dc68abd1f808d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "35fd60a4de76771d907bc996c0e18a6b96bf3e995f769f0807633c879339c459", "generation_seconds_batch": 17.675301842973568, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.055534601211548, "target_z": 0.33203118650304925, "ce": 6.38003396987915, "genuineness": 4.519899368286133, "identity": 0.4764479398727417, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/60858eb98393dad6c71939ddd09e6ef5c6d4afdf565ebb80211dc68abd1f808d.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "6089c1bc1b756fb47c67965c4c3f23985171152a70adec54fd2190ac2c46a1d4", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6640a11ecdd15e883ab06dcd28d283173a473bdf87ed2ce503d0e582a7fb0653", "generation_seconds_batch": 17.77146630291827, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9909073114395142, "target_z": 0.2851562351390115, "ce": 6.260893821716309, "genuineness": 4.457139015197754, "identity": 0.5573132038116455, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6089c1bc1b756fb47c67965c4c3f23985171152a70adec54fd2190ac2c46a1d4.mp3" }, { "id": "heldout-en-0", "language": "en", "sentences": [ "Your message arrived after I had left.", "I read it twice... before putting the phone away.", "Let me know when we can talk." ], "text": "Your message arrived after I had left. I read it twice... before putting the phone away. Let me know when we can talk.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-0", "key": "60a315c932a6fe9a084a369bf917e883f476deb2e4e96b73f1ceb3555fe91298", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Your message arrived after I had left.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Let me know when we can talk.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Your message arrived after I had left.\n(Natural clear speech.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Natural clear speech.) [3.2 seconds duration] Let me know when we can talk.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "00f63c525ecdaada7d30c245b7107ae21d654dc9f274a3200360b6e9a0cdf6aa", "generation_seconds_batch": 23.63622226798907, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.342696189880371, "target_z": 1.2656249837880125, "ce": 5.5659098625183105, "genuineness": 4.306513786315918, "identity": 0.3112753629684448, "hypothesis": "Your message arrived after I had left. I read it twice before putting the phone away. Let me know when we can talk.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/60a315c932a6fe9a084a369bf917e883f476deb2e4e96b73f1ceb3555fe91298.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "60ab9cce5cb715d8743e3572f3126f233ba5fac8f6c445c0111dfa72a3a61b18", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "85c3dea8ac03fcec47cfa240728d49b0eeadb4c26e25acc1cce6e0ef629ab775", "generation_seconds_batch": 21.72275024803821, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7620187997817993, "target_z": 0.11914064188748691, "ce": 6.078212261199951, "genuineness": 4.507347106933594, "identity": 0.6857240200042725, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/60ab9cce5cb715d8743e3572f3126f233ba5fac8f6c445c0111dfa72a3a61b18.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "60db4faffbbe619a1d79cc6c4e95eccfbe590d110115da8e7e8cc7e021dfb3e0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "bb9254202c75ef514f85975bccf28f9cd3cea85126e9115445e32834f8007e40", "generation_seconds_batch": 16.55857592704706, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7089178562164307, "target_z": 1.5312499675760252, "ce": 5.871703147888184, "genuineness": 4.356721878051758, "identity": 0.5983827114105225, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/60db4faffbbe619a1d79cc6c4e95eccfbe590d110115da8e7e8cc7e021dfb3e0.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "617b8ad92b593bd238b98031de946b19ed3acc701e7c0d5ff870a39644a71025", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "fea13235604736d38af4a4342fbe51f2abfd9aa61382c7bf50b878f5c7b76eb7", "generation_seconds_batch": 21.304264158941805, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.647952079772949, "target_z": 0.7617187162250262, "ce": 6.427689552307129, "genuineness": 4.595211505889893, "identity": 0.5804296135902405, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/617b8ad92b593bd238b98031de946b19ed3acc701e7c0d5ff870a39644a71025.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "6221fc5949dca0fd5d080050f453ac68369fe49b45b702f343ac52fc5fdfe348", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1b38ebb47f54233ba2696b053958afc49df1bd4bd3df8ead44bf0d08f27853b2", "generation_seconds_batch": 23.892786954995245, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1928677558898926, "target_z": 0.4316405770395372, "ce": 6.23706579208374, "genuineness": 4.2249250411987305, "identity": 0.40421581268310547, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6221fc5949dca0fd5d080050f453ac68369fe49b45b702f343ac52fc5fdfe348.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "6223b2242aeb2217b7d451d44df588db265f9929e28c51d2b4ea687ac22f2bd9", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "88ca1491b2b260f719cea2eecebb1d745045469ca663bf6f6358379d64d4aadf", "generation_seconds_batch": 17.02645694790408, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.4844465255737305, "target_z": 2.0937502485838073, "ce": 5.4388275146484375, "genuineness": 4.432034492492676, "identity": 0.7471309900283813, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6223b2242aeb2217b7d451d44df588db265f9929e28c51d2b4ea687ac22f2bd9.mp3" }, { "id": "heldout-de-0", "language": "de", "sentences": [ "Deine Nachricht kam, als ich schon weg war.", "Ich las sie zweimal... dann legte ich das Handy weg.", "Sag mir, wann wir reden können." ], "text": "Deine Nachricht kam, als ich schon weg war. Ich las sie zweimal... dann legte ich das Handy weg. Sag mir, wann wir reden können.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-0", "key": "62f7766ca3109000b7d51a1faf0fb935fbed20a24555786d0e55a674b254bb16", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wann wir reden können.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Natural clear speech.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wann wir reden können.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c31dc7beeacaebe11b2bdc98d3291ff9b411730d5a694db5355eeb5d2f35375a", "generation_seconds_batch": 17.012787114013918, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6981465816497803, "target_z": 1.5234374324500524, "ce": 5.6254801750183105, "genuineness": 4.607763767242432, "identity": 0.4873342514038086, "hypothesis": "Deine Nachricht kam, als ich schon weg war. Ich las sie zweimal, dann legte ich das Handy weg. Sag mir, wann wir reden können.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/62f7766ca3109000b7d51a1faf0fb935fbed20a24555786d0e55a674b254bb16.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "6322ec2446a18f19da1236acfebf7c12074a9b6ce09238b09f3aa278a0eb4828", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a0164f7ca7c8e33c7608edf59dbe839e63ef2d01d7737b7379f445102ddd3f09", "generation_seconds_batch": 17.815920863999054, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4719510078430176, "target_z": 1.359375059443954, "ce": 6.133810997009277, "genuineness": 4.871357440948486, "identity": 0.7870618104934692, "hypothesis": "J'ai trouvé la petite boîte, elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6322ec2446a18f19da1236acfebf7c12074a9b6ce09238b09f3aa278a0eb4828.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "6393b43289166b1de412f130d9f02b1c4d884ad73e564e547fa963561d4898b6", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e40c563864189368161aca1a8c8d62ddc5d541663380ce37addb9d79caa3f736", "generation_seconds_batch": 16.6296143559739, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4180946350097656, "target_z": 1.32031238381409, "ce": 6.0305562019348145, "genuineness": 4.337893962860107, "identity": 0.5667842626571655, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6393b43289166b1de412f130d9f02b1c4d884ad73e564e547fa963561d4898b6.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "63d90a08cd693c23cf2f99f970551f755fda234a237f02f02a01e9c8321e72a4", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "104dbbf367ef61185e982f824401ec3ba7af5b588b20cd082af313e97e3e12be", "generation_seconds_batch": 26.507710720063187, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.901075839996338, "target_z": 0.9453124270460566, "ce": 6.23706579208374, "genuineness": 4.325342178344727, "identity": 0.3295360207557678, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/63d90a08cd693c23cf2f99f970551f755fda234a237f02f02a01e9c8321e72a4.mp3" }, { "id": "heldout-fr-1", "language": "fr", "sentences": [ "Les clés sont sur la table de la cuisine.", "J’ai regardé dans le tiroir... puis près de la fenêtre.", "Prends-les quand tu partiras." ], "text": "Les clés sont sur la table de la cuisine. J’ai regardé dans le tiroir... puis près de la fenêtre. Prends-les quand tu partiras.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-1", "key": "642229fa9145f9307ac4afb50db69d9ebb7001b77ed6c4f0525cf67279a1775e", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Prends-les quand tu partiras.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Natural clear speech.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Prends-les quand tu partiras.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "439b780da574c8b84b04e2587e926df6aaf66aad70465aab008f41951d27884c", "generation_seconds_batch": 22.506382621009834, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.838172674179077, "target_z": 1.6250000432319665, "ce": 5.569881439208984, "genuineness": 4.457139015197754, "identity": 0.6737525463104248, "hypothesis": "Les clés sont sur la table de la cuisine. J'ai regardé dans le tiroir, puis près de la fenêtre. Prends-les quand tu partiras.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/642229fa9145f9307ac4afb50db69d9ebb7001b77ed6c4f0525cf67279a1775e.mp3" }, { "id": "heldout-nl-1", "language": "nl", "sentences": [ "De sleutels liggen op de keukentafel.", "Ik keek in de la... en daarna naast het raam.", "Neem ze mee als je weggaat." ], "text": "De sleutels liggen op de keukentafel. Ik keek in de la... en daarna naast het raam. Neem ze mee als je weggaat.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-1", "key": "642322f3dc2efbaec164dbdd0f9c625c6f4d3831b6e27693dc381fcb8c4d192d", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Neem ze mee als je weggaat.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Natural clear speech.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Natural clear speech.) [3.2 seconds duration] Neem ze mee als je weggaat.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7c00e13eb0a968440d05dbd9f8f79abc995a8198acfae7c3dec02de29f0b63a1", "generation_seconds_batch": 16.031477613956667, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.51503586769104, "target_z": 1.3906250270199791, "ce": 5.796247482299805, "genuineness": 4.281409740447998, "identity": 0.4530237317085266, "hypothesis": "De sleutels liggen op de keukentafel. Ik keek in de la en daarna naast het raam. Neem ze mee als je weggaat.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/642322f3dc2efbaec164dbdd0f9c625c6f4d3831b6e27693dc381fcb8c4d192d.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "6432d990458d50ac6194627567109040d8478f11da45eb0da8d38ec199b5e6db", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "ab8c6dbba54468c9d644889cace7ac7bde5c4145bfd1aa1734a141cc4d31a7a7", "generation_seconds_batch": 24.09285667806398, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4396371841430664, "target_z": 1.3359374540660356, "ce": 5.824047088623047, "genuineness": 4.293961524963379, "identity": 0.4676085114479065, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6432d990458d50ac6194627567109040d8478f11da45eb0da8d38ec199b5e6db.mp3" }, { "id": "heldout-fr-0", "language": "fr", "sentences": [ "Ton message est arrivé après mon départ.", "Je l’ai relu... puis j’ai rangé mon téléphone.", "Dis-moi quand nous pourrons parler." ], "text": "Ton message est arrivé après mon départ. Je l’ai relu... puis j’ai rangé mon téléphone. Dis-moi quand nous pourrons parler.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-0", "key": "6464ae80fa254b5e778a44b5f60ad69e9e21a88397e7d178612eeb8ace3be529", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Natural clear speech.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "0e3d37760194f62d4830c4e35a332f31db86b410ce0a55fca8dd1dea09006351", "generation_seconds_batch": 23.048492497066036, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.1720805168151855, "target_z": 1.8671874216420608, "ce": 5.414999485015869, "genuineness": 4.557555198669434, "identity": 0.6878464221954346, "hypothesis": "Ton message is arrived after my départ. Je l'ai relu, puis j'ai rangé mon téléphone. Dis-moi quand nous pourrons parler.", "wer": 0.19047619047619047, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6464ae80fa254b5e778a44b5f60ad69e9e21a88397e7d178612eeb8ace3be529.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "648043be131af00bc3f7953e04448e52a1d4a1bd424c19d7a9f3c1fed4ccf4a2", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "cf6256ace6f93e72b74b11f33e5b6553add32f5320951496ad9c5904bf1b4f4a", "generation_seconds_batch": 24.642582513974048, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0949580669403076, "target_z": 1.0859375405299685, "ce": 5.974957466125488, "genuineness": 4.645420074462891, "identity": 0.5945495367050171, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/648043be131af00bc3f7953e04448e52a1d4a1bd424c19d7a9f3c1fed4ccf4a2.mp3" }, { "id": "heldout-nl-4", "language": "nl", "sentences": [ "Ik schreef het nummer op de envelop.", "Lees het goed... en controleer de laatste twee cijfers.", "Bel me wanneer je even tijd hebt." ], "text": "Ik schreef het nummer op de envelop. Lees het goed... en controleer de laatste twee cijfers. Bel me wanneer je even tijd hebt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-4", "key": "64c146322d8dcc6f55d3aa67236e6709d12dbaad7093dcdec3b31883a13992f7", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Natural clear speech.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Natural clear speech.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "70638bda2c6272507cc52b5a1c2803175ce9b28615490fae181fa494139dbec0", "generation_seconds_batch": 17.41167713503819, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.364238739013672, "target_z": 1.281250054039958, "ce": 5.935244083404541, "genuineness": 4.369274139404297, "identity": 0.29499295353889465, "hypothesis": "Ik schreef het nummer op de envelop. Lees het goed en controleer de laatste twee cijfers. Bel me wanneer je even tijd hebt.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/64c146322d8dcc6f55d3aa67236e6709d12dbaad7093dcdec3b31883a13992f7.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "64e9991759afa7ece970fc382aaadd6e74790562589003ed7579241e2d2c69c5", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "384d7dfce726da354ef7ace1cce362685889241b29d246034bad3cf906d0f5b3", "generation_seconds_batch": 23.864062424981967, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.570615768432617, "target_z": 2.1562498378801256, "ce": 5.478540420532227, "genuineness": 4.595211505889893, "identity": 0.767582893371582, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/64e9991759afa7ece970fc382aaadd6e74790562589003ed7579241e2d2c69c5.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "65142b9e0e0e6de7a8f847800681ba69f73f6524ea88203c11f3e7a74418e710", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "81eb71e424f4a62d5cfa134d015cdeb53fc54055b41f87fa1178aa1f4410a502", "generation_seconds_batch": 23.03462289494928, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6227481365203857, "target_z": 1.4687500324239748, "ce": 6.125868320465088, "genuineness": 4.394378662109375, "identity": 0.4923284947872162, "hypothesis": "I found the little box. It was beside the old window. We can look inside to", "wer": 0.0625, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/65142b9e0e0e6de7a8f847800681ba69f73f6524ea88203c11f3e7a74418e710.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "65e276b052c58263cfbf7a0958bd62e8a3bb94e50bfc3c84089da88ca7875f0d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "5394a41167a1132cb679431341f5b597e8f37a7bb3e07fbd7561558c8ec849cf", "generation_seconds_batch": 25.292252992046997, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.51503586769104, "target_z": 1.3906250270199791, "ce": 6.086154937744141, "genuineness": 4.582659721374512, "identity": 0.46007516980171204, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/65e276b052c58263cfbf7a0958bd62e8a3bb94e50bfc3c84089da88ca7875f0d.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "6681a88cff508cbd67ad4e2a141d0ebf022c5ace5df95243acbb68b507b6bdf3", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ecf254ad28023bc6f8652f20c846fb590099a6d9d873900d1b9108a829d81e56", "generation_seconds_batch": 22.03802254400216, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2888400554656982, "target_z": 1.2265624810860147, "ce": 6.086154937744141, "genuineness": 4.482243061065674, "identity": 0.5355762243270874, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6681a88cff508cbd67ad4e2a141d0ebf022c5ace5df95243acbb68b507b6bdf3.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "675a2d4f94994b8b60b9a07eab77d2e13c43c4b0d293c225f9fc5c8292157e30", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e27ac63f79878e4e7db2f354e5efaf91c569e30bab912273d08bd316d73e3228", "generation_seconds_batch": 23.29884066595696, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7950878143310547, "target_z": 1.5937500756559413, "ce": 5.998785495758057, "genuineness": 4.344170093536377, "identity": 0.40890181064605713, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/675a2d4f94994b8b60b9a07eab77d2e13c43c4b0d293c225f9fc5c8292157e30.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "6769f22c7f43dac57513df902de113398973095afe9369db766c55e4fb944dc0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "56223a730d4eeabaeba0f101265d334f19676206e61758810d6ce5f8599eba24", "generation_seconds_batch": 18.15410049201455, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5365781784057617, "target_z": 1.4062499243440587, "ce": 5.403085231781006, "genuineness": 4.394378662109375, "identity": 0.39320728182792664, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6769f22c7f43dac57513df902de113398973095afe9369db766c55e4fb944dc0.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "67918d46b25b3c268b44bddcea89908317e8fe744704e0ac6ff2f30a9903ec23", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "4cedba1a6f26e6b2abc152e103668582bb29bd1415ecf07620630b8f6841e669", "generation_seconds_batch": 17.6027803060133, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.870486259460449, "target_z": 1.6484374756820188, "ce": 5.982900142669678, "genuineness": 4.381826400756836, "identity": 0.5852096676826477, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam.", "wer": 0.0625, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/67918d46b25b3c268b44bddcea89908317e8fe744704e0ac6ff2f30a9903ec23.mp3" }, { "id": "heldout-nl-2", "language": "nl", "sentences": [ "De weg gaat linksaf na de brug.", "Loop verder... tot aan het kleine houten hek.", "Ik wacht aan de andere kant." ], "text": "De weg gaat linksaf na de brug. Loop verder... tot aan het kleine houten hek. Ik wacht aan de andere kant.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-2", "key": "67e9dd6ce8433f915afcd3b88e4f488a6c00cf534238ef8e58a2da3ec83b2452", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik wacht aan de andere kant.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Natural clear speech.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Natural clear speech.) [3.2 seconds duration] Ik wacht aan de andere kant.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b5fa6c3dfdaf9656cd4c1f62a4767bc6a788c4a630f92e183492218a73233562", "generation_seconds_batch": 14.974486549966969, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9710888862609863, "target_z": 0.9960937324370136, "ce": 5.784333229064941, "genuineness": 4.356721878051758, "identity": 0.45632150769233704, "hypothesis": "De weg gaat linksaf naar de brug. Loop verder tot aan het kleine houten hek. Ik wacht aan de andere kant.", "wer": 0.047619047619047616, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/67e9dd6ce8433f915afcd3b88e4f488a6c00cf534238ef8e58a2da3ec83b2452.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "685efa96eac96d1a195a24400f4c73e193f4a1c62cf435ada4a06fd50a5cff07", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "23796ae56475377aab6106e4b08b51b4c00c26ace7a42b74c93122316cc549e0", "generation_seconds_batch": 24.84561343002133, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.901075839996338, "target_z": 0.9453124270460566, "ce": 5.669164657592773, "genuineness": 4.1998209953308105, "identity": 0.5702605247497559, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás lista.", "wer": 0.06666666666666667, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/685efa96eac96d1a195a24400f4c73e193f4a1c62cf435ada4a06fd50a5cff07.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "6889f34d51c253b0e38ce8e70422787f54aeacbfcd7aef22bc43b22ad8e63d05", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "813e0a45e7d9faf615c7d9952e394019348c822cd5da10b4e1fc4d145aefc299", "generation_seconds_batch": 16.63694982707966, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7843165397644043, "target_z": 1.5859375405299685, "ce": 6.2767791748046875, "genuineness": 4.231201171875, "identity": 0.3564830422401428, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6889f34d51c253b0e38ce8e70422787f54aeacbfcd7aef22bc43b22ad8e63d05.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "68ee80f3734bff2d4ddbfc459d4149d394940b6a847a8fab9385602cb414cdd4", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "648d03c1cee1d75144e9a899316a386e9bfe9410d2529253275646aa62f7a586", "generation_seconds_batch": 17.32473435602151, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2134413719177246, "target_z": 1.1718749081320712, "ce": 6.308549880981445, "genuineness": 4.507347106933594, "identity": 0.3608303666114807, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/68ee80f3734bff2d4ddbfc459d4149d394940b6a847a8fab9385602cb414cdd4.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "691e36d4dba641b1ce695b1ce5a30c42b9df5596b7a82c985e0f958f1ccd8fe6", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "74999e45e46f96bbb4f45bf9286b00dc1290463c24b474da5fd022540ce1f615", "generation_seconds_batch": 23.686055146041326, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6766042709350586, "target_z": 1.5078125351259728, "ce": 5.677107334136963, "genuineness": 4.369274139404297, "identity": 0.567446231842041, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschalten.", "wer": 0.0625, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/691e36d4dba641b1ce695b1ce5a30c42b9df5596b7a82c985e0f958f1ccd8fe6.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "6999eeb67eecf3c2d7baccf695e7c419c7c6f6cf8cbaed35f8f332215eaaff9c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6746d59a3a260c060f68c9a44ad353cc64c3c2a4f9e9a8fc15d75ac6456fc78f", "generation_seconds_batch": 24.499628007994033, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.278069019317627, "target_z": 1.2187501188879077, "ce": 6.2767791748046875, "genuineness": 4.7332844734191895, "identity": 0.6340159177780151, "hypothesis": "He encontrado la cajita. estaba junto a la ventana vieja. podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6999eeb67eecf3c2d7baccf695e7c419c7c6f6cf8cbaed35f8f332215eaaff9c.mp3" }, { "id": "heldout-nl-4", "language": "nl", "sentences": [ "Ik schreef het nummer op de envelop.", "Lees het goed... en controleer de laatste twee cijfers.", "Bel me wanneer je even tijd hebt." ], "text": "Ik schreef het nummer op de envelop. Lees het goed... en controleer de laatste twee cijfers. Bel me wanneer je even tijd hebt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-4", "key": "69d049b5712017b1c0a503833abc820dccd21ddb0691bffb708ca6e932a75ea2", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Natural clear speech.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Natural clear speech.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2796a4b0365cf8f9d51a69e2350152ef0e083e75b23c476ae5320643d8228308", "generation_seconds_batch": 14.909288667025976, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3211536407470703, "target_z": 1.249999913536067, "ce": 5.967014789581299, "genuineness": 4.331617832183838, "identity": 0.5535521507263184, "hypothesis": "Ik schreef het nummer op de envelop. Lees het goed en controleer de laatste twee cijfers. Bel me wanneer je even tijd hebt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/69d049b5712017b1c0a503833abc820dccd21ddb0691bffb708ca6e932a75ea2.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "6a16150344111bcefb810dd8143673a0d20b3eccecaac50dd4d03cfa5c34e56c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "20c532d4e13f052ba22b48ef8eeb0b04208c9941e8311d9b94d4e1f1ffa72c37", "generation_seconds_batch": 23.13790539803449, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.89202880859375, "target_z": 1.6640625459339644, "ce": 5.879645347595215, "genuineness": 4.582659721374512, "identity": 0.5175907015800476, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6a16150344111bcefb810dd8143673a0d20b3eccecaac50dd4d03cfa5c34e56c.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "6a5a321d3273f5ca18bd75904b3da09953093cd81eef7bc38f04c908da34fa66", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "5bc02eaa34a1327898cfb1f67e6b54b3355348970b2415f8119a8c94e578199e", "generation_seconds_batch": 22.414825026062317, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.053597450256348, "target_z": 1.781250226967824, "ce": 5.831989288330078, "genuineness": 4.758388519287109, "identity": 0.7496635913848877, "hypothesis": "J'ai trouvé la petite boîte, elle était près de la vieille fenêtre, nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6a5a321d3273f5ca18bd75904b3da09953093cd81eef7bc38f04c908da34fa66.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "6a603493e71b02d46e5bb8250bf802781469afde68313a4e4ec00d1b3b57e01f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "f48a140c5f2865e5331149b0bb4647b655f08199c082698685d5ba8b1e2e1c1a", "generation_seconds_batch": 22.03919477108866, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.450408458709717, "target_z": 1.3437499891920084, "ce": 5.9431867599487305, "genuineness": 4.369274139404297, "identity": 0.5377486944198608, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6a603493e71b02d46e5bb8250bf802781469afde68313a4e4ec00d1b3b57e01f.mp3" }, { "id": "heldout-en-4", "language": "en", "sentences": [ "I wrote the number on the envelope.", "Read it carefully... and check the last two digits.", "Call me when you have a moment." ], "text": "I wrote the number on the envelope. Read it carefully... and check the last two digits. Call me when you have a moment.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-4", "key": "6a61bc20d43b3e97640d8dd45682f7055c5dbaa5aa987cdb55f8a7d7a432a2c2", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I wrote the number on the envelope.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Call me when you have a moment.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I wrote the number on the envelope.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Call me when you have a moment.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "fad0fbff76da2be15aeda87d46678302c7efcd40574a22356375818181d9996a", "generation_seconds_batch": 17.358769585960545, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2625031471252441, "target_z": -0.2431640749967403, "ce": 6.356205940246582, "genuineness": 4.017816066741943, "identity": 0.3708779215812683, "hypothesis": "I wrote the number on the envelope, read it carefully, and check the last two digits. Call me when you have a moment.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6a61bc20d43b3e97640d8dd45682f7055c5dbaa5aa987cdb55f8a7d7a432a2c2.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "6a920ddd8764f32d0df4a8f6e631d80e85cbdac7f132238ca5a267c73f19a9e0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "41312d05422d13f6be3d0273921381ff93b1c41bdbef3c29bd191d5d1210c60a", "generation_seconds_batch": 24.03318282496184, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8976278901100159, "target_z": -0.5078124918940062, "ce": 5.9431867599487305, "genuineness": 4.457139015197754, "identity": 0.5999182462692261, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6a920ddd8764f32d0df4a8f6e631d80e85cbdac7f132238ca5a267c73f19a9e0.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "6b1f7386d7486d2fb17d8e0733e984cb9fde603f8efeb1b053621106a438e110", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0a5e091ebde2ab02c5227489365361105deef0cba1c6e01a3df8ffd72e78446f", "generation_seconds_batch": 24.70089547301177, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.938775062561035, "target_z": 0.9726561270590953, "ce": 5.974957466125488, "genuineness": 4.262581825256348, "identity": 0.4356754720211029, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6b1f7386d7486d2fb17d8e0733e984cb9fde603f8efeb1b053621106a438e110.mp3" }, { "id": "heldout-es-0", "language": "es", "sentences": [ "Tu mensaje llegó después de que me fuera.", "Lo leí dos veces... y guardé el teléfono.", "Dime cuándo podemos hablar." ], "text": "Tu mensaje llegó después de que me fuera. Lo leí dos veces... y guardé el teléfono. Dime cuándo podemos hablar.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-0", "key": "6b87091cdfa68d268e30ffb87b44cfd7e8844815de4dffbe95423bd192621e2f", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo podemos hablar.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Natural clear speech.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo podemos hablar.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "ea09ab3d8ef8289fdd7a4e1a6a0d45f41dba7213e9ecc3e7a52be24cf6416271", "generation_seconds_batch": 14.724913107929751, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5796632766723633, "target_z": 1.4375000648479497, "ce": 5.728734970092773, "genuineness": 4.432034492492676, "identity": 0.3464667797088623, "hypothesis": "Tu mensaje llegó después de que me fuera, lo leí dos veces y guardé el teléfono. Dime cuándo podemos hablar.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6b87091cdfa68d268e30ffb87b44cfd7e8844815de4dffbe95423bd192621e2f.mp3" }, { "id": "heldout-es-3", "language": "es", "sentences": [ "Alguien ha movido la pequeña silla azul.", "Estaba aquí... junto a la estantería alta.", "La colocaremos después de comer." ], "text": "Alguien ha movido la pequeña silla azul. Estaba aquí... junto a la estantería alta. La colocaremos después de comer.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-3", "key": "6b973c750e3616c70f28663c29496a0c74bea4a87dff4bad65cadc5394b99131", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La colocaremos después de comer.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La colocaremos después de comer.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "cf067aec3e03e7c6e3b44d44e33b32758102ddac938a5df32acd7536e117d690", "generation_seconds_batch": 17.430664474959485, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2476927042007446, "target_z": -0.2539062675629864, "ce": 6.292664527893066, "genuineness": 4.6203155517578125, "identity": 0.6765201091766357, "hypothesis": "Alguien ha movido la pequeña silla azul. Estaba aquí, junto a la estantería alta. La colocaremos después de comer.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6b973c750e3616c70f28663c29496a0c74bea4a87dff4bad65cadc5394b99131.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "6bb74122d0c3912105cb42d0738f6e5410d2d26e623643aa1f94f55590350c0e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5da472c53de8544f1253b700d9a68b06d144efd81e24746bd66232e0b5028454", "generation_seconds_batch": 24.642582513974048, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4217562675476074, "target_z": 0.5976561702910618, "ce": 6.221180438995361, "genuineness": 4.65797233581543, "identity": 0.5510914921760559, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6bb74122d0c3912105cb42d0738f6e5410d2d26e623643aa1f94f55590350c0e.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "6bea9332cd9687d3288814e518bfe7a310f5f00fa450b92a3e8c717361cfe4e9", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5df0d0b89451d563106cd21ec839d934bd293fc68d35053a9733cb35fad34a38", "generation_seconds_batch": 17.058061757939868, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.202670097351074, "target_z": 1.1640623730060984, "ce": 6.403861999511719, "genuineness": 4.507347106933594, "identity": 0.6674965023994446, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6bea9332cd9687d3288814e518bfe7a310f5f00fa450b92a3e8c717361cfe4e9.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "6beb8856f6804551c0c7c961825937727efd18c2b4f8d4c2eee462647a6fe14b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c4d901e324b389a8ecfcbc5e8c65242064e700a21d509e77e6d8a05bbd25bfe7", "generation_seconds_batch": 23.049930661916733, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.238645553588867, "target_z": 0.4648437648609885, "ce": 6.38003396987915, "genuineness": 4.406930446624756, "identity": 0.7327660322189331, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6beb8856f6804551c0c7c961825937727efd18c2b4f8d4c2eee462647a6fe14b.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "6cae4229bc6159ad29db9b6030ade67054f3e45d9d08e335789978975c5003cd", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "31f29fc44d3272d10e96224b92ff240db72d99c38c94e8443e1335a1fae791f6", "generation_seconds_batch": 22.144952556001954, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7089178562164307, "target_z": 1.5312499675760252, "ce": 5.784333229064941, "genuineness": 4.268857479095459, "identity": 0.26209479570388794, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6cae4229bc6159ad29db9b6030ade67054f3e45d9d08e335789978975c5003cd.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "6d8f8947f2f5ce4a289ff7f233d074dc2a0413ac978db5f3a28632bca3df07c3", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "885743aed61a9ffc258719429da182f0f5ac188acfa1cbaaad70d4fd7e6e42b0", "generation_seconds_batch": 24.499628007994033, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4217562675476074, "target_z": 0.5976561702910618, "ce": 6.006728172302246, "genuineness": 4.494794845581055, "identity": 0.8056857585906982, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6d8f8947f2f5ce4a289ff7f233d074dc2a0413ac978db5f3a28632bca3df07c3.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "6dc79932c7f067a3dc46357efacbe261d7226defc3bedb18a9e50b2f6d459887", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f8c300c209397c9a48cf0b67d87483bfb6e6a8f946c13146e10918823327f2bb", "generation_seconds_batch": 25.080359576968476, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.364238739013672, "target_z": 1.281250054039958, "ce": 6.109982967376709, "genuineness": 4.344170093536377, "identity": 0.41367989778518677, "hypothesis": "I found the little box. It was beside the old window. We can look inside to", "wer": 0.0625, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6dc79932c7f067a3dc46357efacbe261d7226defc3bedb18a9e50b2f6d459887.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "6e3583d6acc172b0112bd79797119f466525411a2f6a271d1d6893b6f6b10e55", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9e961bcbe2ccbc544a06a88712bceb33eac7de53be489ea70e84e51805dea6d3", "generation_seconds_batch": 15.610294860904105, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0303306579589844, "target_z": 1.039062502701998, "ce": 6.403861999511719, "genuineness": 4.519899368286133, "identity": 0.5501828789710999, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6e3583d6acc172b0112bd79797119f466525411a2f6a271d1d6893b6f6b10e55.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "6e88c1dfd83ca6f777e512c1ed28b603f2831480c98130955f08c2bcf7d99616", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f88bf5c74001b372a6899352ee6c36d158e6bd95415b4a57e48b04bb7cb6d1c3", "generation_seconds_batch": 23.202125952928327, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3857812881469727, "target_z": 1.2968751242919037, "ce": 6.117925643920898, "genuineness": 4.821148872375488, "identity": 0.532371997833252, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6e88c1dfd83ca6f777e512c1ed28b603f2831480c98130955f08c2bcf7d99616.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "6eb88cc8e3b37208d6429b84fbdf0563af0216d376046d1be136c026a1a63a7b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1ffe66091d87988bc6ae2eaf6c9d36f41466c0bffe9d9f57af38cc7fb4320372", "generation_seconds_batch": 21.72275024803821, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4119539260864258, "target_z": -0.1347656256754995, "ce": 6.165581703186035, "genuineness": 4.419482707977295, "identity": 0.547112762928009, "hypothesis": "ik heb het kleine doosje gevonden, het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6eb88cc8e3b37208d6429b84fbdf0563af0216d376046d1be136c026a1a63a7b.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "6f21eea14bb2dfa677a82675c52c4ff34b1cf62cab61016d2a0dab6f145a0a37", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "01bbc59d0224667caec2800fb3f0a01cddf9b101033d714ad5ed92317b04fd4f", "generation_seconds_batch": 24.34498886798974, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.773545265197754, "target_z": 1.578125005403996, "ce": 5.780362129211426, "genuineness": 4.519899368286133, "identity": 0.5015958547592163, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6f21eea14bb2dfa677a82675c52c4ff34b1cf62cab61016d2a0dab6f145a0a37.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "6f599421d365d0dc880e3d8d51e78ded5c7b6dda7accf90414a912e83663ccc5", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "825aec6861b97fdbe73751472ed3e94081b1b28c9b5e38b0706c4d200e807e62", "generation_seconds_batch": 22.618776715011336, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2996113300323486, "target_z": 1.2343750162119875, "ce": 5.736677646636963, "genuineness": 4.1998209953308105, "identity": 0.5411572456359863, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6f599421d365d0dc880e3d8d51e78ded5c7b6dda7accf90414a912e83663ccc5.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "6faefb66694e6a9978acb60b8cc20bfcbc0528aeabe246ef7883448ee8eaf39d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a40aff5dac7b508b90d5b1684eca6be11ac36442bdf8d3c03e4988076f87ddcc", "generation_seconds_batch": 24.47754904802423, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.933389663696289, "target_z": 0.9687500324239748, "ce": 6.252951145172119, "genuineness": 4.432034492492676, "identity": 0.8450807929039001, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/6faefb66694e6a9978acb60b8cc20bfcbc0528aeabe246ef7883448ee8eaf39d.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "7028d23b59e96fa9d78ca5d8a989bdb5257a1cd06043ff22f25c572170ad5079", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f6f6c6aa8938aa47cfc6e6852a5fbceab7ea4c3638870e9d9ded0cdc286a0147", "generation_seconds_batch": 16.75300578493625, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6119768619537354, "target_z": 1.460937497298002, "ce": 5.1052350997924805, "genuineness": 4.143336772918701, "identity": 0.4241582751274109, "hypothesis": "La puerta sigue abierta. He dejado tu abrijo sobre la silla. Dime cuándo estás lista.", "wer": 0.13333333333333333, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7028d23b59e96fa9d78ca5d8a989bdb5257a1cd06043ff22f25c572170ad5079.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "706c780430d181ff3f7d10cfb44dd8b2cbb26fc2b973dbcbe883effc88230817", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "386ecc3dccb97d583a288eed548b43b8cd0f78a818237311d6cc0e004da7f94b", "generation_seconds_batch": 23.864062424981967, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.301335334777832, "target_z": 1.960937497298002, "ce": 5.6413655281066895, "genuineness": 4.356721878051758, "identity": 0.8413527011871338, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/706c780430d181ff3f7d10cfb44dd8b2cbb26fc2b973dbcbe883effc88230817.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "70831e17495eaa42b3a214af15e8ffcccb989e81b6d96c31e0894dca396633fc", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b3aaec00b4f83a0d99fba4d1306c8c55c83c622c8d4cacc252ed47eb0a94cfbf", "generation_seconds_batch": 24.70089547301177, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.437913179397583, "target_z": 0.609374972980021, "ce": 5.927301406860352, "genuineness": 3.8922953605651855, "identity": 0.43830862641334534, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/70831e17495eaa42b3a214af15e8ffcccb989e81b6d96c31e0894dca396633fc.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "70c998f294595ca726bc71fdfcb928fef18e37c77f3a9f8112d0f18f5c093379", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "97c7561dce8db0d257a192a3ed28fc778dd07ec7798794952db487b11cb2a8ec", "generation_seconds_batch": 23.613548352965154, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.419818878173828, "target_z": 2.0468750378279705, "ce": 5.748591423034668, "genuineness": 4.444586753845215, "identity": 0.41058990359306335, "hypothesis": "He encontrado la cajita, estaba junto a la ventana vieja, podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/70c998f294595ca726bc71fdfcb928fef18e37c77f3a9f8112d0f18f5c093379.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "711b2663284beeeb4434ee5ae15d64972db65863e08f1935c52e7ced1f3b6bd1", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "5890a99c9011fe04122829e2afef75f5664b7450d41c9abccee71dbf102ca618", "generation_seconds_batch": 17.863600790966302, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9889698028564453, "target_z": 1.7343750162119875, "ce": 6.062326908111572, "genuineness": 4.6203155517578125, "identity": 0.7750499248504639, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/711b2663284beeeb4434ee5ae15d64972db65863e08f1935c52e7ced1f3b6bd1.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "7130fb865cf413a40d9a99892b9eb3de82e805a6e9f207bdeaba1de1013e9ae7", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f41572281a97ff7038ebf1c5b3673c8944c3391b25b1a3bf373f10ec578381eb", "generation_seconds_batch": 24.547520775930025, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.568892002105713, "target_z": 1.4296875297219769, "ce": 6.1020402908325195, "genuineness": 4.545003414154053, "identity": 0.8088749051094055, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7130fb865cf413a40d9a99892b9eb3de82e805a6e9f207bdeaba1de1013e9ae7.mp3" }, { "id": "heldout-nl-1", "language": "nl", "sentences": [ "De sleutels liggen op de keukentafel.", "Ik keek in de la... en daarna naast het raam.", "Neem ze mee als je weggaat." ], "text": "De sleutels liggen op de keukentafel. Ik keek in de la... en daarna naast het raam. Neem ze mee als je weggaat.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-1", "key": "71d920fe830770e359bbef49683c634915f346a18e304a54279e38156147849c", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Neem ze mee als je weggaat.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Natural clear speech.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Natural clear speech.) [3.2 seconds duration] Neem ze mee als je weggaat.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2d8b479581b7f6d7d0a70bf5eb238c5c5669f3775c792fd50df281ba15ee545f", "generation_seconds_batch": 25.929826888954267, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5796632766723633, "target_z": 1.4375000648479497, "ce": 5.77639102935791, "genuineness": 4.344170093536377, "identity": 0.6010235548019409, "hypothesis": "De sleutels liggen op de keukentafel. Ik keek in de la en daarna, naast de traam, neem ze mee als je weggaat.", "wer": 0.09090909090909091, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/71d920fe830770e359bbef49683c634915f346a18e304a54279e38156147849c.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "72807e90403010c310e0aaa45b59d0e4e036ec6865e78abb921213942890aa4a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f148624c7f2687bac25e14b138197c6a4687de5332a0f1c55aa695611fab1edb", "generation_seconds_batch": 16.797331872046925, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6442904472351074, "target_z": 1.4843749297480544, "ce": 5.863759994506836, "genuineness": 4.306513786315918, "identity": 0.4263330101966858, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/72807e90403010c310e0aaa45b59d0e4e036ec6865e78abb921213942890aa4a.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "7297f26ca8e3a478984cbdd8aad6db8fcea298d9a7d13033def240d9ce87ce3b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "185063c6ebdb6636bd8839461b7faeb979c844f552e13315cc9e9a4a573b3f4b", "generation_seconds_batch": 24.330811747000553, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7395076751708984, "target_z": 0.8281250918679287, "ce": 5.982900142669678, "genuineness": 4.432034492492676, "identity": 0.5939849615097046, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7297f26ca8e3a478984cbdd8aad6db8fcea298d9a7d13033def240d9ce87ce3b.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "72e1fc3562ae8ef034d3e5f67a1fd53ae68d40729169929ea93fd3fa8e4616fe", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3b28381211a82966beadf0303652447ef1502d5de9324ce7b8c59234ee669039", "generation_seconds_batch": 24.083597186021507, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9997410774230957, "target_z": 1.74218755133796, "ce": 6.022613525390625, "genuineness": 4.65797233581543, "identity": 0.2443220317363739, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás lista.", "wer": 0.06666666666666667, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/72e1fc3562ae8ef034d3e5f67a1fd53ae68d40729169929ea93fd3fa8e4616fe.mp3" }, { "id": "heldout-nl-4", "language": "nl", "sentences": [ "Ik schreef het nummer op de envelop.", "Lees het goed... en controleer de laatste twee cijfers.", "Bel me wanneer je even tijd hebt." ], "text": "Ik schreef het nummer op de envelop. Lees het goed... en controleer de laatste twee cijfers. Bel me wanneer je even tijd hebt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-4", "key": "72e6d066fdc28f026868f16a382efdf956680f9379bb117a57648080724480a3", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Natural clear speech.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Natural clear speech.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "88640469b7fd5c885ba968f4466366886b4d307bf770ea876b1606ff340d8f11", "generation_seconds_batch": 24.535028933081776, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2349839210510254, "target_z": 1.1874999783840168, "ce": 5.744620323181152, "genuineness": 4.231201171875, "identity": 0.4376160204410553, "hypothesis": "Ik schreef het nummer op de envelop. Lees het goed en controleer de laatste twee cijfers. Bel me wanneer je even tijd hebt.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/72e6d066fdc28f026868f16a382efdf956680f9379bb117a57648080724480a3.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "7325bc46599496db20d49eff98638ac933bb48a2bd699f473809c151256e2d13", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "cd6aef14d08f19d375cb8332ee0075fc1858668f436b2fb4782c98969f201750", "generation_seconds_batch": 25.28117792296689, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.51503586769104, "target_z": 1.3906250270199791, "ce": 5.839932441711426, "genuineness": 4.432034492492676, "identity": 0.5301257371902466, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7325bc46599496db20d49eff98638ac933bb48a2bd699f473809c151256e2d13.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "73bd6a7dd926b36d3465dba98243b4b207c40b7982d2255cd9fb90353b29224e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f519e04decba52a52ec8dff9f218e1ba01a1230ae99212d214e68aa647f0c322", "generation_seconds_batch": 24.140849750023335, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.701808214187622, "target_z": 0.8007812189270241, "ce": 6.252951145172119, "genuineness": 4.394378662109375, "identity": 0.6454854011535645, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/73bd6a7dd926b36d3465dba98243b4b207c40b7982d2255cd9fb90353b29224e.mp3" }, { "id": "heldout-nl-2", "language": "nl", "sentences": [ "De weg gaat linksaf na de brug.", "Loop verder... tot aan het kleine houten hek.", "Ik wacht aan de andere kant." ], "text": "De weg gaat linksaf na de brug. Loop verder... tot aan het kleine houten hek. Ik wacht aan de andere kant.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-2", "key": "748afb920964120a47669bb2ffb4bff766af6468f5d95be53c35b8e338d54a6d", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik wacht aan de andere kant.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Natural clear speech.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Natural clear speech.) [3.2 seconds duration] Ik wacht aan de andere kant.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8e86af15d3b858b93ab98de8e012e86a47547954728bc26cdd1ee17ea4dc8a42", "generation_seconds_batch": 17.09770359203685, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6981465816497803, "target_z": 1.5234374324500524, "ce": 6.197352409362793, "genuineness": 4.68307638168335, "identity": 0.17183183133602142, "hypothesis": "De weg gaat linksaf naar de brug. Loop verder tot aan het kleine, houten hek. Ik wacht aan de andere kant.", "wer": 0.047619047619047616, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/748afb920964120a47669bb2ffb4bff766af6468f5d95be53c35b8e338d54a6d.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "74d60a752e9a8ea02b1d7ba3c7294e1eedfcc55caa769c11290ce83f527f549a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ae89c714a96f3b968a3c6401b49949c474949cbbf23ae84c13eadb5a044791f6", "generation_seconds_batch": 25.812747161951847, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5348541736602783, "target_z": 0.679687443258044, "ce": 6.252951145172119, "genuineness": 4.7458367347717285, "identity": 0.6006615161895752, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/74d60a752e9a8ea02b1d7ba3c7294e1eedfcc55caa769c11290ce83f527f549a.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "74f41b5052bfffcfcacf14db7c9988e0b5da81fd6707f7873f409f33108f8e15", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "dd15c1d14cad04e165f7a826a7a92c7f09af8af1bdcbb2eeda19fb0bc8d5d7ee", "generation_seconds_batch": 25.080359576968476, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.051873207092285, "target_z": 1.0546875729539433, "ce": 5.879645347595215, "genuineness": 4.1998209953308105, "identity": 0.5422084331512451, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/74f41b5052bfffcfcacf14db7c9988e0b5da81fd6707f7873f409f33108f8e15.mp3" }, { "id": "heldout-de-0", "language": "de", "sentences": [ "Deine Nachricht kam, als ich schon weg war.", "Ich las sie zweimal... dann legte ich das Handy weg.", "Sag mir, wann wir reden können." ], "text": "Deine Nachricht kam, als ich schon weg war. Ich las sie zweimal... dann legte ich das Handy weg. Sag mir, wann wir reden können.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-0", "key": "757d02d1ad1a927ace4cad1f925a06e8c7e61845c74b14160fe70d18f11e08dc", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wann wir reden können.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Natural clear speech.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wann wir reden können.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4940cbd1930c5ac64fd1331df7371da26dee93bd5804818a80652f3aab49f8e8", "generation_seconds_batch": 23.707801511045545, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.687375545501709, "target_z": 1.5156250702519456, "ce": 5.764476776123047, "genuineness": 4.607763767242432, "identity": 0.4484926462173462, "hypothesis": "Deine Nachricht kam, als ich schon weg war. Ich las sie zweimal, dann legte ich das Handy weg. Sag mir, wann wir reden können.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/757d02d1ad1a927ace4cad1f925a06e8c7e61845c74b14160fe70d18f11e08dc.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "757e43070df18093491bac3c703b6c18d3d1f13d7717945ca262091df388125d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e1f95a44472cecf0940f71dd4fd9e830b8eaba88711d006b4be3179b4c46851f", "generation_seconds_batch": 17.327590133994818, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.342696189880371, "target_z": 1.2656249837880125, "ce": 6.229123115539551, "genuineness": 4.7081804275512695, "identity": 0.49675098061561584, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/757e43070df18093491bac3c703b6c18d3d1f13d7717945ca262091df388125d.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "75918193a7ab77aea3a69911bc99b5ba77df7fb4c2f36d1308fc1fa69bb742d5", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c667fcabb6994966b6f0a60d2a962ff894535e063a2dc56a1ac346af4ec6abaa", "generation_seconds_batch": 23.261335554067045, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7297050952911377, "target_z": 0.09570312297350157, "ce": 6.165581703186035, "genuineness": 4.582659721374512, "identity": 0.5930361747741699, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/75918193a7ab77aea3a69911bc99b5ba77df7fb4c2f36d1308fc1fa69bb742d5.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "759d0b7c507c4e33d6fd661b86ef4f903245c31237fae80a853005e53830faa5", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "07913514b1d3af3ec2f45d6daddc80884232347acb6a9590972ee5f20ba44d55", "generation_seconds_batch": 23.68443738296628, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.750278949737549, "target_z": 0.8359376269939015, "ce": 6.23706579208374, "genuineness": 4.557555198669434, "identity": 0.7752597332000732, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/759d0b7c507c4e33d6fd661b86ef4f903245c31237fae80a853005e53830faa5.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "75ba7c29c4861e3dd2a38b2b1f60a52453ed4b4017c893572b36aa0998596e8b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1d905e94624c23b587b3133439216e0d64dab233ba2ff249ef31122c00de1e3a", "generation_seconds_batch": 25.4117253680015, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0178353786468506, "target_z": 0.3046874864900105, "ce": 5.919358730316162, "genuineness": 4.570107460021973, "identity": 0.5912840366363525, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/75ba7c29c4861e3dd2a38b2b1f60a52453ed4b4017c893572b36aa0998596e8b.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "76266b66c6ff0cd253ade9dea0426bdcbf4cc4b31627dd78297fdd67d9c0ecfa", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b4dce0b793948512ee831182ca195fcd8689d4ecd5a53c8faceebd3f1eccdc92", "generation_seconds_batch": 15.919662689091638, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4719510078430176, "target_z": 1.359375059443954, "ce": 6.149696350097656, "genuineness": 4.632867813110352, "identity": 0.6038088798522949, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/76266b66c6ff0cd253ade9dea0426bdcbf4cc4b31627dd78297fdd67d9c0ecfa.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "768dbed23ef02b0e1bdee2d810c20c040fba7c4a0d752d08fd389a453489c739", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ccdf64a7cb191c2dff1b9445567140dffe6ee1a3e4b6e937de0008eddaa10e34", "generation_seconds_batch": 24.92359939101152, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0770771503448486, "target_z": 0.3476562567549948, "ce": 6.3641486167907715, "genuineness": 4.532451152801514, "identity": 0.5681760311126709, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/768dbed23ef02b0e1bdee2d810c20c040fba7c4a0d752d08fd389a453489c739.mp3" }, { "id": "heldout-fr-2", "language": "fr", "sentences": [ "La route tourne à gauche après le pont.", "Continue... jusqu’à la petite barrière en bois.", "Je t’attendrai de l’autre côté." ], "text": "La route tourne à gauche après le pont. Continue... jusqu’à la petite barrière en bois. Je t’attendrai de l’autre côté.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-2", "key": "76b91de54b1735303f1d380bcb7d89912e35173e4163613b76dbeda90506a0d6", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Natural clear speech.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Natural clear speech.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "026d7935b165166649a4ce575c039a1edf0fd377201803c30103b993a00a0b20", "generation_seconds_batch": 15.972894505946897, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.945884943008423, "target_z": 1.7031250486359624, "ce": 5.847874641418457, "genuineness": 4.482243061065674, "identity": 0.7485866546630859, "hypothesis": "La route tourne à gauche après le pont, continue jusqu'à la petite barrière en bois, je t'attendrai de l'autre côté.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/76b91de54b1735303f1d380bcb7d89912e35173e4163613b76dbeda90506a0d6.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "771e1e75433367916298b089963fcc6c951e09109d7ed315ddad3d40b7c73529", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "678e20fc66ed6157db17946dcc3a948be02c8875f4feb745bb8adb5e9b85720d", "generation_seconds_batch": 23.556757037062198, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.752002716064453, "target_z": 1.5624999351520503, "ce": 5.792276382446289, "genuineness": 4.469690799713135, "identity": 0.24203315377235413, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/771e1e75433367916298b089963fcc6c951e09109d7ed315ddad3d40b7c73529.mp3" }, { "id": "heldout-nl-1", "language": "nl", "sentences": [ "De sleutels liggen op de keukentafel.", "Ik keek in de la... en daarna naast het raam.", "Neem ze mee als je weggaat." ], "text": "De sleutels liggen op de keukentafel. Ik keek in de la... en daarna naast het raam. Neem ze mee als je weggaat.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-1", "key": "776082c9dfea2aa758e005b276cc0b30fffa5e374b5b8484ef2e0ecb3f32f9e7", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Neem ze mee als je weggaat.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Natural clear speech.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Natural clear speech.) [3.2 seconds duration] Neem ze mee als je weggaat.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c6153db528ee0a6ac203369a189cd6a20feeb4208836c95c1b146940635eb499", "generation_seconds_batch": 16.031477613956667, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0626444816589355, "target_z": 1.0625001080799161, "ce": 6.157639026641846, "genuineness": 4.519899368286133, "identity": 0.6983897089958191, "hypothesis": "De sleutels liggen op de keukentafel. Ik keek in de la en daarna naast het raam neem ze mee als je weggaat.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/776082c9dfea2aa758e005b276cc0b30fffa5e374b5b8484ef2e0ecb3f32f9e7.mp3" }, { "id": "heldout-fr-3", "language": "fr", "sentences": [ "Quelqu’un a déplacé la petite chaise bleue.", "Elle était ici... près de la grande bibliothèque.", "Nous la remettrons après le déjeuner." ], "text": "Quelqu’un a déplacé la petite chaise bleue. Elle était ici... près de la grande bibliothèque. Nous la remettrons après le déjeuner.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-3", "key": "77962898fb95ddc0e92a153772191fe01482e40f98d2245158cb92652e8a93c1", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8bcf7f14a18960045484e2644b729ebd0a205bda13932e1668c1b0dab61d1105", "generation_seconds_batch": 17.520918608061038, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4486844539642334, "target_z": 0.6171875081059938, "ce": 5.990842819213867, "genuineness": 4.6203155517578125, "identity": 0.8238123655319214, "hypothesis": "Quelqu'un a déplacé la petite chaise bleue. Elle était ici, près de la grande bibliothèque. Nous la remettrons après le déjeuner.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/77962898fb95ddc0e92a153772191fe01482e40f98d2245158cb92652e8a93c1.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "77a4013519d88fbe73c2916540aa013562f5d5c7caef56cbad944afb33fbd628", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "ed0647b1ba75965f317609e9d9e65dbdf8eb598b4c2a46b962508494b7f68365", "generation_seconds_batch": 17.815920863999054, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.245755195617676, "target_z": 1.1953125135099896, "ce": 6.205295085906982, "genuineness": 4.457139015197754, "identity": 0.8015519380569458, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/77a4013519d88fbe73c2916540aa013562f5d5c7caef56cbad944afb33fbd628.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "77a62d2e0d4d33e04f62b9e00ee03a6ebc622529a847a4eb5a714b80ccfe3fda", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "535dfd0a40ef5e1cfe1534412b1d09f055e3b9072dcb8948e10f8c0495c3c631", "generation_seconds_batch": 22.957556341891177, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.114776372909546, "target_z": 0.37499995676803355, "ce": 6.427689552307129, "genuineness": 4.595211505889893, "identity": 0.5882214307785034, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/77a62d2e0d4d33e04f62b9e00ee03a6ebc622529a847a4eb5a714b80ccfe3fda.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "77e8a4c98ae818d5bb8d4f207601caea7da1464c087314b4e9b26f3b5eff428c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e38d0bda110bd9e560a0e8594cd4d651821046723a028831e29de1584fe1280a", "generation_seconds_batch": 24.146093792049214, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.191899061203003, "target_z": 1.1562500108079916, "ce": 5.879645347595215, "genuineness": 4.256305694580078, "identity": 0.6064823865890503, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/77e8a4c98ae818d5bb8d4f207601caea7da1464c087314b4e9b26f3b5eff428c.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "784557618646bdac735ac80b4a3340b37a8faac8532f700f9703e0f553b11f99", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "e0da3885ad2ddde58f32a7f562da26ca9e14cb3b4e3eb58903200a385149e74d", "generation_seconds_batch": 24.333596550975926, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.247479438781738, "target_z": 1.92187516752387, "ce": 5.9431867599487305, "genuineness": 4.469690799713135, "identity": 0.8337043523788452, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/784557618646bdac735ac80b4a3340b37a8faac8532f700f9703e0f553b11f99.mp3" }, { "id": "heldout-nl-2", "language": "nl", "sentences": [ "De weg gaat linksaf na de brug.", "Loop verder... tot aan het kleine houten hek.", "Ik wacht aan de andere kant." ], "text": "De weg gaat linksaf na de brug. Loop verder... tot aan het kleine houten hek. Ik wacht aan de andere kant.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-2", "key": "788dcdf1408fc166ee61ed82b033dfd3580d86f3eeda111ba341b8c561e4af97", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik wacht aan de andere kant.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Natural clear speech.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Natural clear speech.) [3.2 seconds duration] Ik wacht aan de andere kant.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ebb8607006eec04919cf533ec38370ac5fc69bea7217fcc9dc6b81ce06fd3684", "generation_seconds_batch": 24.60657593992073, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.019559383392334, "target_z": 1.0312499675760252, "ce": 5.617537498474121, "genuineness": 4.319066047668457, "identity": 0.26573893427848816, "hypothesis": "De weg gaat linksaf naar de brug. Loop verder tot aan het kleine houten hek. Ik wacht aan de andere kant.", "wer": 0.047619047619047616, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/788dcdf1408fc166ee61ed82b033dfd3580d86f3eeda111ba341b8c561e4af97.mp3" }, { "id": "heldout-fr-2", "language": "fr", "sentences": [ "La route tourne à gauche après le pont.", "Continue... jusqu’à la petite barrière en bois.", "Je t’attendrai de l’autre côté." ], "text": "La route tourne à gauche après le pont. Continue... jusqu’à la petite barrière en bois. Je t’attendrai de l’autre côté.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-2", "key": "78b7a1da321486de0cb0076aa94fb25434c99b2df0cfaceaac980909d45f9f3c", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Natural clear speech.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Natural clear speech.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "871673805f0fa07d91682c27fd9fa5a5c3f88d52ce1eb867dff42fe33b24ab23", "generation_seconds_batch": 15.972894505946897, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.859714984893799, "target_z": 1.640624940556046, "ce": 5.577824115753174, "genuineness": 4.545003414154053, "identity": 0.8218933343887329, "hypothesis": "La route tourne à gauche après le pont, continue jusqu'à la petite barrière en bois, je t'attendrai de l'autre côté.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/78b7a1da321486de0cb0076aa94fb25434c99b2df0cfaceaac980909d45f9f3c.mp3" }, { "id": "heldout-nl-1", "language": "nl", "sentences": [ "De sleutels liggen op de keukentafel.", "Ik keek in de la... en daarna naast het raam.", "Neem ze mee als je weggaat." ], "text": "De sleutels liggen op de keukentafel. Ik keek in de la... en daarna naast het raam. Neem ze mee als je weggaat.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-1", "key": "794d3b9296e3c4be7c1e4182a2fa3cf095f7d593990763ce0f19c1f5e499a989", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Neem ze mee als je weggaat.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Neem ze mee als je weggaat.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3357a1485f89bd248ada6a1837c8f4c4d67610cd6793bb8937836d3d66d98cba", "generation_seconds_batch": 17.59364196297247, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2975096702575684, "target_z": -0.2177734223012618, "ce": 6.308549880981445, "genuineness": 4.607763767242432, "identity": 0.5028052926063538, "hypothesis": "De sleutels liggen op de keukentafel. Ik keek in de la en daarna naast het raam neem ze mee als je weggaat.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/794d3b9296e3c4be7c1e4182a2fa3cf095f7d593990763ce0f19c1f5e499a989.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "7a289164919f297111ff8e71bd079c6f8c7029d8ef9a9f8cc57dca2f9ca81e72", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3aa33de0ef409c017acb22ae4125997eb6ad4659bdd9f689532fb82bb8fd01ab", "generation_seconds_batch": 18.084932073019445, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.376733779907227, "target_z": 2.01562489732408, "ce": 5.967014789581299, "genuineness": 4.6956281661987305, "identity": 0.6756267547607422, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi, quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7a289164919f297111ff8e71bd079c6f8c7029d8ef9a9f8cc57dca2f9ca81e72.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "7aa9b8946c6fc1781ecbe1089f6f47f14fc205f18e78420468f5952a3b3bbbc3", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5cd7c1f6d79ea66e77addda550b9deca7bb07f23fbf97b6ff07017b36db7bd8b", "generation_seconds_batch": 23.223652250948362, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.1720805168151855, "target_z": 1.8671874216420608, "ce": 5.77639102935791, "genuineness": 4.394378662109375, "identity": 0.7160022854804993, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7aa9b8946c6fc1781ecbe1089f6f47f14fc205f18e78420468f5952a3b3bbbc3.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "7ab2d3149d5092af8bf527c36d7e087f1e04e4364cbeeb127ea59a2765d8d6ac", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9e567e935434119c0ac719d9233d22450cc91cccde9c2fd12b7b5993fc2eeec6", "generation_seconds_batch": 25.363981387927197, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6766042709350586, "target_z": 1.5078125351259728, "ce": 5.847874641418457, "genuineness": 4.256305694580078, "identity": 0.2653052806854248, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7ab2d3149d5092af8bf527c36d7e087f1e04e4364cbeeb127ea59a2765d8d6ac.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "7ac3cf23683604334fa85b84c105961030a06a6685db33d18bcc43761dff4116", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3d4069ef9efa25af87c6b935c18cebdfb00d90906fc3547a1f20be1b6a3843c6", "generation_seconds_batch": 17.16528386902064, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.723350763320923, "target_z": 0.8164062891789696, "ce": 6.133810997009277, "genuineness": 4.055472373962402, "identity": 0.3835373818874359, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7ac3cf23683604334fa85b84c105961030a06a6685db33d18bcc43761dff4116.mp3" }, { "id": "heldout-nl-4", "language": "nl", "sentences": [ "Ik schreef het nummer op de envelop.", "Lees het goed... en controleer de laatste twee cijfers.", "Bel me wanneer je even tijd hebt." ], "text": "Ik schreef het nummer op de envelop. Lees het goed... en controleer de laatste twee cijfers. Bel me wanneer je even tijd hebt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-4", "key": "7ae11347b4faf064a968688b017d711978b765e1f36f2e60d9129107dae17ae2", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Natural clear speech.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Natural clear speech.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e4bd731fd32d8cff54621ae7da563cb2fd86b2138830f90525ec981824184569", "generation_seconds_batch": 17.41167713503819, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.202670097351074, "target_z": 1.1640623730060984, "ce": 6.221180438995361, "genuineness": 4.457139015197754, "identity": 0.6951421499252319, "hypothesis": "Ik schreef het nummer op de envelop. Lees het goed en controleer de laatste twee cijfers. Bel me wanneer je even tijd hebt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7ae11347b4faf064a968688b017d711978b765e1f36f2e60d9129107dae17ae2.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "7bdc482e59962e16ebc1125e5fae8c05b5c7f114be4f8f8373c976ab019c6008", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4aba1db1f902fc10bc49090b7ab0fac99f44ddbdc8d0e2ce7e3eda368641c6a7", "generation_seconds_batch": 23.610983759048395, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5365781784057617, "target_z": 1.4062499243440587, "ce": 6.086154937744141, "genuineness": 4.149612903594971, "identity": 0.3235897719860077, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7bdc482e59962e16ebc1125e5fae8c05b5c7f114be4f8f8373c976ab019c6008.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "7c0851749bdcf65a6ffc6536ca4bea3f87491566809f3b9140371bfb18efd8bb", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4572afca7e35b278cc115c222442219c6223245a84d197fb6808ed7b28a5df6c", "generation_seconds_batch": 23.431125467061065, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4934933185577393, "target_z": 1.3749999567680335, "ce": 5.510311126708984, "genuineness": 4.231201171875, "identity": 0.5144003629684448, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7c0851749bdcf65a6ffc6536ca4bea3f87491566809f3b9140371bfb18efd8bb.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "7cd840e3e3121c542f77f97ccd58297cb40a997a398a58f59d74d69aa03e35a3", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "76395f9e66a36785acfe91aaf9e199bdf1cf412d3538bdd94e3e940163985dd8", "generation_seconds_batch": 22.83592540200334, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6533377170562744, "target_z": 0.7656249837880126, "ce": 6.300607204437256, "genuineness": 4.532451152801514, "identity": 0.5530966520309448, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7cd840e3e3121c542f77f97ccd58297cb40a997a398a58f59d74d69aa03e35a3.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "7d2147ea52a2bc97336c6144116ec01f62ec19088618e7b27daf78cfa92acf4f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6b66521676e75b97fdf1ad5c7045fdae466e4b2c4cc6c08cec74923b893b5a00", "generation_seconds_batch": 23.879537864006124, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.562751054763794, "target_z": -0.025390652695478525, "ce": 6.2767791748046875, "genuineness": 4.519899368286133, "identity": 0.5579500794410706, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7d2147ea52a2bc97336c6144116ec01f62ec19088618e7b27daf78cfa92acf4f.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "7d3a09eca17b315400ab74d17c71c21c7aa003689f6299cd2fe5172d590df1a0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "022b6f12c0f0ed938872d472ae2ae7c20d742a3ac20ca78ba179d441f600409e", "generation_seconds_batch": 27.454596178024076, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.859714984893799, "target_z": 1.640624940556046, "ce": 5.8875885009765625, "genuineness": 4.494794845581055, "identity": 0.40389958024024963, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7d3a09eca17b315400ab74d17c71c21c7aa003689f6299cd2fe5172d590df1a0.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "7d82de716078accb165ab38d8db5ef491236bcc1c3fc3a3333b4562e5165220a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0998896743758a10dd75a34c206b1c5b413b4bf9fd46823f37e12154a3c5f873", "generation_seconds_batch": 23.730586632038467, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.547349452972412, "target_z": 1.4140624594700315, "ce": 5.839932441711426, "genuineness": 4.105680465698242, "identity": 0.24493855237960815, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7d82de716078accb165ab38d8db5ef491236bcc1c3fc3a3333b4562e5165220a.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "7de0011102957df3909bd601ce9e4d123ba0915a2d5feeeb7257dcc653d2db69", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "756accc45ed2d69a3034a8829431852aa535720ea92b2311e38651484586bf72", "generation_seconds_batch": 27.22625131707173, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.010512351989746, "target_z": 1.750000086463933, "ce": 5.637393951416016, "genuineness": 4.444586753845215, "identity": 0.49883925914764404, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que.", "wer": 0.11764705882352941, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7de0011102957df3909bd601ce9e4d123ba0915a2d5feeeb7257dcc653d2db69.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "7e1c08d4f49843320298aa24ada2353e5e795b83e8069234143abf5585df8818", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f0c0f6d205b7076b914b337da3cd8ff1dd2373cc8706dc6b66e82c738861b99d", "generation_seconds_batch": 23.33019120595418, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.719689130783081, "target_z": 1.539062502701998, "ce": 6.133810997009277, "genuineness": 4.507347106933594, "identity": 0.4899441599845886, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7e1c08d4f49843320298aa24ada2353e5e795b83e8069234143abf5585df8818.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "7e259194e81f1510a9e82907982fcf43ae690aaca3e53967438817525440480a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "bb2527946effa3b3dfc0c276072941dad0d65b61cd1efdb4c949496c0aa6157d", "generation_seconds_batch": 15.236449437914416, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.301335334777832, "target_z": 1.960937497298002, "ce": 5.681078910827637, "genuineness": 4.494794845581055, "identity": 0.6626132726669312, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7e259194e81f1510a9e82907982fcf43ae690aaca3e53967438817525440480a.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "7ec1e5728ce1daa57bef4217b15e8e45f5e62af81ddb98773fb4ed9d35616e54", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d2fcc1749433c182dd0d20820f5d11b39086fd999ac7a5c9cefc5d1005ccc41d", "generation_seconds_batch": 24.610660327016376, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.987245798110962, "target_z": 1.0078125351259728, "ce": 5.863759994506836, "genuineness": 4.2374773025512695, "identity": 0.36750978231430054, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7ec1e5728ce1daa57bef4217b15e8e45f5e62af81ddb98773fb4ed9d35616e54.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "7ec2d5f5341b27d7f6916469e96a6f94f078dec1f35c8f5fea2b3b8fedf41c82", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5ebe22b30a7336e53832c7eca8af519dbebd56a61f80502364393d08247d95d6", "generation_seconds_batch": 23.172279790975153, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.547349452972412, "target_z": 1.4140624594700315, "ce": 5.597680568695068, "genuineness": 4.356721878051758, "identity": 0.6478216648101807, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7ec2d5f5341b27d7f6916469e96a6f94f078dec1f35c8f5fea2b3b8fedf41c82.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "7f00c568d86df9da452b93cfedaa69a891e95b3f3bbb82de40f5c87a3d8e3db1", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6772568a6d6966c8664e56a629d717bdfd62982fd07217b1bb23899048d971a6", "generation_seconds_batch": 15.919662689091638, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.256526470184326, "target_z": 1.2031250486359624, "ce": 6.1020402908325195, "genuineness": 4.457139015197754, "identity": 0.522815465927124, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7f00c568d86df9da452b93cfedaa69a891e95b3f3bbb82de40f5c87a3d8e3db1.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "7f1525ef20ff0e0b636d93761048c868cca5425ccadfa47cc2e4b82189df656a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "231b1c134f87bb1f08a1c96ce3af82a8b06b6da6944b45ed338c22bd27f61c29", "generation_seconds_batch": 20.989678737008944, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6856513023376465, "target_z": 0.789062416238065, "ce": 6.181467056274414, "genuineness": 4.231201171875, "identity": 0.3028170168399811, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7f1525ef20ff0e0b636d93761048c868cca5425ccadfa47cc2e4b82189df656a.mp3" }, { "id": "heldout-en-1", "language": "en", "sentences": [ "The keys are on the kitchen table.", "I checked the drawer... then looked beside the window.", "Take them with you when you leave." ], "text": "The keys are on the kitchen table. I checked the drawer... then looked beside the window. Take them with you when you leave.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-1", "key": "7f6fce1a6f733719f45a363f28fee4e24b2f07a3396aa07b42392913d7c24239", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The keys are on the kitchen table.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Take them with you when you leave.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The keys are on the kitchen table.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Take them with you when you leave.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9d19c67376f1e7609827e218315e78130c801af1b2303937eff13bb2582b25dd", "generation_seconds_batch": 15.52945992606692, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5708295106887817, "target_z": -0.01953125135099895, "ce": 6.300607204437256, "genuineness": 4.444586753845215, "identity": 0.4219222962856293, "hypothesis": "The keys are on the kitchen table. I checked the drawer, then looked beside the window. Take them with you when you leave.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7f6fce1a6f733719f45a363f28fee4e24b2f07a3396aa07b42392913d7c24239.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "7f71bb5c268a82023fcec9d1d7a05bfa636f95a309fa29a86145516abdaed7a7", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c3d1223e0a3397aecb95d796d444b0fff4c177a34d849e2561e19c4604d00e76", "generation_seconds_batch": 24.495948707102798, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.547349452972412, "target_z": 1.4140624594700315, "ce": 5.871703147888184, "genuineness": 4.287685871124268, "identity": 0.5308353304862976, "hypothesis": "La puerta sigue abierta, he dejado tu abrigo sobre la silla. Dime cuándo estás lista.", "wer": 0.06666666666666667, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7f71bb5c268a82023fcec9d1d7a05bfa636f95a309fa29a86145516abdaed7a7.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "7f84405af62aac010d033feb1ad9a08acafce725cd99e3e3ff095986566a24aa", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e8c70a34837037d03cd309a8159fc7b04c539263673a49131abf8ba1f2a2da3a", "generation_seconds_batch": 24.543410731013864, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.031299352645874, "target_z": 0.31445306893354347, "ce": 5.752562999725342, "genuineness": 3.8420867919921875, "identity": 0.5971933603286743, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás lista.", "wer": 0.06666666666666667, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7f84405af62aac010d033feb1ad9a08acafce725cd99e3e3ff095986566a24aa.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "7fab78a5c3ec0f4e40ec5fe43ef1a2fbbaa78aeaf3f63d360db53a92c09e1e9c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5d00cb57a58a9e65a10a3cffaa4503dce0f1c15dc96229bbae5f1ac7b7a1ddd4", "generation_seconds_batch": 23.542010444914922, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.364238739013672, "target_z": 1.281250054039958, "ce": 6.165581703186035, "genuineness": 4.582659721374512, "identity": 0.7180138230323792, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/7fab78a5c3ec0f4e40ec5fe43ef1a2fbbaa78aeaf3f63d360db53a92c09e1e9c.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "805092699a8173f4f89e2fbd346e7e32bbd77609d79be60be5ab02894042bdbe", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "4922db83dc077f4f0244afe2c97918e0f5ad2ad38710c3733ec9b27be5ac262c", "generation_seconds_batch": 17.474135274998844, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.450408458709717, "target_z": 1.3437499891920084, "ce": 5.71682071685791, "genuineness": 4.20609712600708, "identity": 0.2933942973613739, "hypothesis": "I found the little box. It was beside the old window. We can look inside.", "wer": 0.0625, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/805092699a8173f4f89e2fbd346e7e32bbd77609d79be60be5ab02894042bdbe.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "80a22a56d66e24d6dbfb183216894fedad3b4baf5c1b8dff6723e1edc375e868", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d0434b95697da195a4573da7b9ca93f22a6478f94f3bc56c73e7899f3b5735d2", "generation_seconds_batch": 23.882941075018607, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.342696189880371, "target_z": 1.2656249837880125, "ce": 6.2767791748046875, "genuineness": 4.306513786315918, "identity": 0.4825771450996399, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/80a22a56d66e24d6dbfb183216894fedad3b4baf5c1b8dff6723e1edc375e868.mp3" }, { "id": "heldout-es-1", "language": "es", "sentences": [ "Las llaves están en la mesa de la cocina.", "Miré en el cajón... y luego junto a la ventana.", "Llévatelas cuando te vayas." ], "text": "Las llaves están en la mesa de la cocina. Miré en el cajón... y luego junto a la ventana. Llévatelas cuando te vayas.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-1", "key": "80c2597a6b8f43b5338a6b37d654868d679f46ded48514b42837ab18fba327ad", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Llévatelas cuando te vayas.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Llévatelas cuando te vayas.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "978753ebc708c7f654ee855f6517152dec7174c39c08f41a62ccb999515b4251", "generation_seconds_batch": 17.944635855033994, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7579796314239502, "target_z": 0.1162109844472136, "ce": 6.252951145172119, "genuineness": 4.457139015197754, "identity": 0.6877202391624451, "hypothesis": "Las llaves están en la mesa de la cocina. Miré en el cajón y luego junto a la ventana. Llévatelas cuando te vayas.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/80c2597a6b8f43b5338a6b37d654868d679f46ded48514b42837ab18fba327ad.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "8108bf53888b971b09434884db9a46b21ad1bc6e3526f271094e632c98d963bb", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4480c4c3dbc4361c88ef04d0cb3db410111c2a75a37ebfed4193fcaaa4d09650", "generation_seconds_batch": 24.888466550968587, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.07513952255249, "target_z": 1.7968749513640376, "ce": 6.173524379730225, "genuineness": 4.413206577301025, "identity": 0.6769927740097046, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8108bf53888b971b09434884db9a46b21ad1bc6e3526f271094e632c98d963bb.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "8138da63b0abf06877346be8eb9a87eedd410bc645848073eeccf67651cbcd40", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "166e32687bd3301c951c556365a45f8c717361628c1d6ed9ea9505eecf361c4d", "generation_seconds_batch": 24.610660327016376, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.601205587387085, "target_z": 1.4531249621720292, "ce": 6.2767791748046875, "genuineness": 4.231201171875, "identity": 0.27247217297554016, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8138da63b0abf06877346be8eb9a87eedd410bc645848073eeccf67651cbcd40.mp3" }, { "id": "heldout-en-2", "language": "en", "sentences": [ "The road turns left behind the bridge.", "Keep going... until you reach the small wooden gate.", "I will meet you on the other side." ], "text": "The road turns left behind the bridge. Keep going... until you reach the small wooden gate. I will meet you on the other side.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-2", "key": "81526f424a94089e01563a80ea072c834a0fb48cde8d2b273e1eaf1f195ef2f7", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The road turns left behind the bridge.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I will meet you on the other side.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The road turns left behind the bridge.\n(Natural clear speech.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Natural clear speech.) [3.2 seconds duration] I will meet you on the other side.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "d06756c933cade8a25ec34b8080dad76a3c8209d1c57e9bc7e07d17dee8db2e7", "generation_seconds_batch": 17.88071251299698, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9764745235443115, "target_z": 1.0, "ce": 6.197352409362793, "genuineness": 4.256305694580078, "identity": 0.3218165338039398, "hypothesis": "The road turns left behind the bridge. Keep going until you reach the small wooden gate. I will meet you on the other side.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/81526f424a94089e01563a80ea072c834a0fb48cde8d2b273e1eaf1f195ef2f7.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "82968872daf6cf52517490edc8193dfa66fa64405283e18b8ab34592f96ba108", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7974eabe3b25eaa207d4c6aa3f7724ce0e09d2205651c96144226522da00e4f4", "generation_seconds_batch": 22.102732184925117, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.118224143981934, "target_z": 1.828124746012197, "ce": 6.070269584655762, "genuineness": 4.557555198669434, "identity": 0.8341007232666016, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/82968872daf6cf52517490edc8193dfa66fa64405283e18b8ab34592f96ba108.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "82e1fd825f44b7446a212b09bf7554c9a9e6fea6e6a3ddcca75dfc9b9f55abe4", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "92f3ab4f51c30522dee3b4ad966618d6dea8a03937405919b3f5d39c14ea6efc", "generation_seconds_batch": 23.598842017934658, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.031299352645874, "target_z": 0.31445306893354347, "ce": 6.252951145172119, "genuineness": 4.444586753845215, "identity": 0.5036823749542236, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/82e1fd825f44b7446a212b09bf7554c9a9e6fea6e6a3ddcca75dfc9b9f55abe4.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "82fe5545aefe799cba6dccdc3a8473ee05ad148ed8b5c635ee91e13608eef620", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "b5c2320284210d0d9705d3dcfdcfb4b5cff22d0fe3be28d8e6fa18a2be144978", "generation_seconds_batch": 22.144952556001954, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.752002716064453, "target_z": 1.5624999351520503, "ce": 5.871703147888184, "genuineness": 4.394378662109375, "identity": 0.012676313519477844, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/82fe5545aefe799cba6dccdc3a8473ee05ad148ed8b5c635ee91e13608eef620.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "835cca84fe8cd3e5a4a52a388bf464773fdd02d104ed05f6c9a63f9ab75cd26d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "093c2c6dda677e83a69d9f7ec1aaa4f9868bef168876211543c372969f6d6541", "generation_seconds_batch": 26.599157833959907, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9351136684417725, "target_z": 1.6953125135099896, "ce": 5.824047088623047, "genuineness": 4.432034492492676, "identity": 0.565858006477356, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hinein.", "wer": 0.0625, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/835cca84fe8cd3e5a4a52a388bf464773fdd02d104ed05f6c9a63f9ab75cd26d.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "838765f306b07d22300b935c2e5bf28e3abfbf1fb4d719608202a8f088a0f4aa", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2647cf9d609ebfb75d3bf38af53edd66b7a5e4c12abd29609c47e4ae9385965a", "generation_seconds_batch": 22.008515903027728, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.105729103088379, "target_z": 1.0937499027280755, "ce": 6.078212261199951, "genuineness": 4.6203155517578125, "identity": 0.49878036975860596, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/838765f306b07d22300b935c2e5bf28e3abfbf1fb4d719608202a8f088a0f4aa.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "83e0346f803e77b1f8dc6a0744efe26af3589a58bc3a2dd06687024c994ee140", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6bff45c0003796258fe0837e603c7a0bd1117f87fccd6190456b64364d1f57d1", "generation_seconds_batch": 22.24144284101203, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5581207275390625, "target_z": 1.421874994596004, "ce": 6.117925643920898, "genuineness": 4.570107460021973, "identity": 0.7468422055244446, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/83e0346f803e77b1f8dc6a0744efe26af3589a58bc3a2dd06687024c994ee140.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "8450e9c40c90e58edb4a8128c66feb2ba92452e8886371cc16aaa5c61a44449c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6088fbe3bc2c0564f37154834514b7dc23a734e4c7dca31e8af1360909797275", "generation_seconds_batch": 24.05071882903576, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3625144958496094, "target_z": 0.5546874000260775, "ce": 6.213237762451172, "genuineness": 3.9048471450805664, "identity": 0.3105505108833313, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8450e9c40c90e58edb4a8128c66feb2ba92452e8886371cc16aaa5c61a44449c.mp3" }, { "id": "heldout-es-4", "language": "es", "sentences": [ "Escribí el número en el sobre.", "Léelo bien... y comprueba las dos últimas cifras.", "Llámame cuando tengas un momento." ], "text": "Escribí el número en el sobre. Léelo bien... y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-4", "key": "84aa77d2cec8a3f90508b0047211556627da8583d48b8e6267d7e7a8f1b3cbe6", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Escribí el número en el sobre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Llámame cuando tengas un momento.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Escribí el número en el sobre.\n(Natural clear speech.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Natural clear speech.) [3.2 seconds duration] Llámame cuando tengas un momento.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "423dc048e53359b2ca172d9f9d32bd7d7565a42ea58c9bedbbf937c9ae7cfaad", "generation_seconds_batch": 15.019184255041182, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.51503586769104, "target_z": 1.3906250270199791, "ce": 5.577824115753174, "genuineness": 4.369274139404297, "identity": 0.5554066896438599, "hypothesis": "Escribí el número en el sobre. Lílo bien y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "wer": 0.05263157894736842, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/84aa77d2cec8a3f90508b0047211556627da8583d48b8e6267d7e7a8f1b3cbe6.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "84aa7b94ed6574398aed57070c0210c756bf0c2713d847db13bb22ff47bf2a77", "bucket": "vn_S_ASMR_reg", "phase": "wording", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ebbe2a3c0a205dfb2cd657beae8be528551ee83b54422dda2d3de45638fb10dd", "generation_seconds_batch": 16.535874380962923, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8903045654296875, "target_z": 0.9374998919200838, "ce": 6.2767791748046875, "genuineness": 4.093128204345703, "identity": 0.217843160033226, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/84aa7b94ed6574398aed57070c0210c756bf0c2713d847db13bb22ff47bf2a77.mp3" }, { "id": "heldout-es-2", "language": "es", "sentences": [ "La carretera gira a la izquierda tras el puente.", "Sigue... hasta la pequeña puerta de madera.", "Te esperaré al otro lado." ], "text": "La carretera gira a la izquierda tras el puente. Sigue... hasta la pequeña puerta de madera. Te esperaré al otro lado.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-2", "key": "84d97f7ba5c37e6b44303a2e6054c275db4d7f7f34c2a2b0a030f08abb26985d", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Te esperaré al otro lado.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Natural clear speech.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Natural clear speech.) [3.2 seconds duration] Te esperaré al otro lado.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "eb9edb4e434f6647115b7a3145c4df8ea86857810804c30607ce52d8d9e23eda", "generation_seconds_batch": 23.485097950091586, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4180946350097656, "target_z": 1.32031238381409, "ce": 5.367343425750732, "genuineness": 4.400654315948486, "identity": 0.6235189437866211, "hypothesis": "La carretera gira a la izquierda tras el puente. Sigue hasta la pequeña puerta de madera. Te esperaré al otro lado.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/84d97f7ba5c37e6b44303a2e6054c275db4d7f7f34c2a2b0a030f08abb26985d.mp3" }, { "id": "heldout-fr-4", "language": "fr", "sentences": [ "J’ai écrit le numéro sur l’enveloppe.", "Lis-le bien... et vérifie les deux derniers chiffres.", "Appelle-moi quand tu auras un moment." ], "text": "J’ai écrit le numéro sur l’enveloppe. Lis-le bien... et vérifie les deux derniers chiffres. Appelle-moi quand tu auras un moment.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-4", "key": "854e484e49526cb8305c24bd91b83b74b712329e0751990e71252b5b1f4327bb", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Natural clear speech.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Natural clear speech.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5ad97b663fec684c74ae4e7efbe5a1c823058ea9183511b84da24fb7b5ea2889", "generation_seconds_batch": 17.38643721397966, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.687375545501709, "target_z": 1.5156250702519456, "ce": 5.764476776123047, "genuineness": 4.645420074462891, "identity": 0.7485928535461426, "hypothesis": "J'ai écrit le numéro sur l'enveloppe, lis le bien et vérifie les deux derniers chiffres. Appelle-moi quand tu auras un moment.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/854e484e49526cb8305c24bd91b83b74b712329e0751990e71252b5b1f4327bb.mp3" }, { "id": "heldout-fr-1", "language": "fr", "sentences": [ "Les clés sont sur la table de la cuisine.", "J’ai regardé dans le tiroir... puis près de la fenêtre.", "Prends-les quand tu partiras." ], "text": "Les clés sont sur la table de la cuisine. J’ai regardé dans le tiroir... puis près de la fenêtre. Prends-les quand tu partiras.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-1", "key": "85631e8cb971fda4cf0c24756572e89b36f81f78ef057cf53d8858c267308bf7", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Prends-les quand tu partiras.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Natural clear speech.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Prends-les quand tu partiras.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b26a4b3e7077be9ec59c6d7dff835136c9fe4bceefb6b414faf7358445fdf732", "generation_seconds_batch": 18.414642314077355, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8166301250457764, "target_z": 1.6093749729800209, "ce": 5.855817794799805, "genuineness": 4.432034492492676, "identity": 0.6998509168624878, "hypothesis": "Les clés sont sur la table de la cuisine. J'ai regardé dans le tiroir, puis près de la fenêtre. Prends-les quand tu partiras.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/85631e8cb971fda4cf0c24756572e89b36f81f78ef057cf53d8858c267308bf7.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "85cf3663cc87f851e7ba23b1802a1de8db6e620e002ddb6b981d51ab9a93887e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "30ea83c336c582e5725bc3da478737312cc1535a8efbec7f0af3a35b5de29b76", "generation_seconds_batch": 23.008963654050604, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.568892002105713, "target_z": 1.4296875297219769, "ce": 5.7684478759765625, "genuineness": 4.532451152801514, "identity": 0.6221628189086914, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay detrás.", "wer": 0.058823529411764705, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/85cf3663cc87f851e7ba23b1802a1de8db6e620e002ddb6b981d51ab9a93887e.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "85ec7113d604220928512381178667a3201c50bb13adbe65db1957c144fd4597", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8bf4787502ce0bc9a5c26fb6b6885b67abab7c7d32681263443dbde7fbfef52e", "generation_seconds_batch": 23.116177136078477, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9135711193084717, "target_z": 1.679687443258044, "ce": 6.173524379730225, "genuineness": 4.68307638168335, "identity": 0.46520906686782837, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/85ec7113d604220928512381178667a3201c50bb13adbe65db1957c144fd4597.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "860ee13a51456ac87c16b51e0744649172c23533200a5a214762d2f7ca28ab6c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "22176b8d5de16421a6888fd3f1000b2ee9ce697c9c475054310a049c3fcf6ba9", "generation_seconds_batch": 24.687736685969867, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9351136684417725, "target_z": 1.6953125135099896, "ce": 5.871703147888184, "genuineness": 4.306513786315918, "identity": 0.8353784084320068, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/860ee13a51456ac87c16b51e0744649172c23533200a5a214762d2f7ca28ab6c.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "867fead2d3c48c974dc18288b73a44ee6bd756d847c08cea5e2df92c9830cdef", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6bad2007c5460f140ca115b9cbab20d76484e40ec168074a06b7b090a317c0d4", "generation_seconds_batch": 22.811438572010957, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1272716522216797, "target_z": 1.1093749729800209, "ce": 6.078212261199951, "genuineness": 4.3504462242126465, "identity": 0.6198369264602661, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/867fead2d3c48c974dc18288b73a44ee6bd756d847c08cea5e2df92c9830cdef.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "86849120cf4277750201f5783a17ce0e5e2810b123fff6e2748fb822ccecbb52", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a29cecd4500f7d29cc798c36d84fd22ae8cdc6ade560d25198851bd691d99725", "generation_seconds_batch": 17.671789115993306, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.667770504951477, "target_z": 0.05078121892702409, "ce": 6.046441555023193, "genuineness": 4.545003414154053, "identity": 0.5401096940040588, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/86849120cf4277750201f5783a17ce0e5e2810b123fff6e2748fb822ccecbb52.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "873ff27559cfdd0d25b49f8980e4d05d8847bf5b38208ddace5286fdd234011a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "f2982ae6f70287f62141796fd4e1b8281a2074c948652d547531b3e59b6ded8d", "generation_seconds_batch": 24.463273369008675, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4934933185577393, "target_z": 1.3749999567680335, "ce": 5.998785495758057, "genuineness": 4.557555198669434, "identity": 0.7954223155975342, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/873ff27559cfdd0d25b49f8980e4d05d8847bf5b38208ddace5286fdd234011a.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "8746411d7327cebd70d56352396a2f3a8460597c7be9e4081df01e83879ab003", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3f3c720390f7746f16f830e80c05bf3a1cdd318b0dd70070f84d88f11a45d447", "generation_seconds_batch": 24.547843346023, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.355191230773926, "target_z": 1.9999998270721342, "ce": 5.375286102294922, "genuineness": 4.582659721374512, "identity": 0.7168443202972412, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8746411d7327cebd70d56352396a2f3a8460597c7be9e4081df01e83879ab003.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "87cd43ef758e9039a3f6222bb4bc84bf66e32f3107a5d3a5f2211d4be4a2c128", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "81bf249a51ac218f6a1ad7888d4fc5acfea2c18da422e27ba4e22928599bbe44", "generation_seconds_batch": 17.40729280503001, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3177056312561035, "target_z": -0.20312504863596229, "ce": 6.173524379730225, "genuineness": 4.457139015197754, "identity": 0.769963264465332, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/87cd43ef758e9039a3f6222bb4bc84bf66e32f3107a5d3a5f2211d4be4a2c128.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "87e280d5729f6428e68f0e6c22d5a2c7061f298d16f371e971e4a603648e158d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3a6194c274c08c5346c343fc7692f07bb90c0df0ef4550f4bf9420a184e6a56d", "generation_seconds_batch": 24.98371786705684, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.247479438781738, "target_z": 1.92187516752387, "ce": 5.95112943649292, "genuineness": 4.545003414154053, "identity": 0.7119519710540771, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/87e280d5729f6428e68f0e6c22d5a2c7061f298d16f371e971e4a603648e158d.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "883882d352759451ee236627de39b0d0fc385b6192a87dae61c171491f484f05", "bucket": "vn_S_ASMR_reg", "phase": "wording", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3aaa3233f4f6dd8e415677454a593b2df666d1e20d86f1b6ef0eb3038c43d8eb", "generation_seconds_batch": 17.095417739939876, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.497154951095581, "target_z": 0.6523437432450052, "ce": 6.292664527893066, "genuineness": 4.532451152801514, "identity": 0.6192329525947571, "hypothesis": "Die Tür ist doch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.05263157894736842, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/883882d352759451ee236627de39b0d0fc385b6192a87dae61c171491f484f05.mp3" }, { "id": "heldout-es-4", "language": "es", "sentences": [ "Escribí el número en el sobre.", "Léelo bien... y comprueba las dos últimas cifras.", "Llámame cuando tengas un momento." ], "text": "Escribí el número en el sobre. Léelo bien... y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-4", "key": "88b5c1231df7c55ef2bd18526743aa0e3893bc31659ea369841bb2458587d3a9", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Escribí el número en el sobre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Llámame cuando tengas un momento.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Escribí el número en el sobre.\n(Natural clear speech.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Natural clear speech.) [3.2 seconds duration] Llámame cuando tengas un momento.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "269d691818010d6dba4c28bdb718eb941b2e6ef890c5277c65c31a448738d867", "generation_seconds_batch": 15.019184255041182, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.396552324295044, "target_z": 1.3046874864900104, "ce": 5.816103935241699, "genuineness": 4.394378662109375, "identity": 0.4904666543006897, "hypothesis": "Escribí el número en el sobre. Léelo bien y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/88b5c1231df7c55ef2bd18526743aa0e3893bc31659ea369841bb2458587d3a9.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "88d39a7ac61d81981756a0e6630b5f5c9bb35e7c8e5735202a65adb135bb33ba", "bucket": "vn_S_ASMR_reg", "phase": "wording", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "cdbfa588b313cd7943c6a0529d9ca8da1aedf996913aeb32498c139fa1e704b4", "generation_seconds_batch": 15.99711354705505, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2548022270202637, "target_z": 0.4765623946220817, "ce": 5.879645347595215, "genuineness": 4.243753433227539, "identity": 0.6037828922271729, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/88d39a7ac61d81981756a0e6630b5f5c9bb35e7c8e5735202a65adb135bb33ba.mp3" }, { "id": "heldout-es-2", "language": "es", "sentences": [ "La carretera gira a la izquierda tras el puente.", "Sigue... hasta la pequeña puerta de madera.", "Te esperaré al otro lado." ], "text": "La carretera gira a la izquierda tras el puente. Sigue... hasta la pequeña puerta de madera. Te esperaré al otro lado.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-2", "key": "896ff04573e9570b1f8af881f73c833cd51fb44a9d354e5cf7458c753fc00959", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Te esperaré al otro lado.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Natural clear speech.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Natural clear speech.) [3.2 seconds duration] Te esperaré al otro lado.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c82d122f952c83d3845087ba7a389229c62e842827f75567a01cb9a594ceb02c", "generation_seconds_batch": 17.201339381048456, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.396552324295044, "target_z": 1.3046874864900104, "ce": 6.109982967376709, "genuineness": 4.557555198669434, "identity": 0.5917937159538269, "hypothesis": "La carretera gira a la izquierda tras el puente. Sigue hasta la pequeña puerta de madera. Te esperaré al otro lado.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/896ff04573e9570b1f8af881f73c833cd51fb44a9d354e5cf7458c753fc00959.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "89cac3b24ad53b1bec1904486cfab9ad5c9f47f2b9a6dff34239b3478e4ef8e7", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0b5c0a4c3b383b4c4f2f00c6816bf3f1a734b56e0e653d7e463176882255e85b", "generation_seconds_batch": 24.548229896929115, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0087881088256836, "target_z": 1.0234374324500524, "ce": 5.736677646636963, "genuineness": 4.118232727050781, "identity": 0.33937323093414307, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/89cac3b24ad53b1bec1904486cfab9ad5c9f47f2b9a6dff34239b3478e4ef8e7.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "89f2eb321df0d91bf2d31124360b336d1ac4f912dfad856ccfa3db142c273681", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 131, "seconds_generated": 10.48, "mp3_sha256": "8d6376fe8161885a24e28306854ca59fc8964e4c900f5d49049bb0a9a4650a0c", "generation_seconds_batch": 25.275906038004905, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8489439487457275, "target_z": 1.6328125783579392, "ce": 5.49045467376709, "genuineness": 4.319066047668457, "identity": 0.4021362066268921, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás.", "wer": 0.06666666666666667, "duration_error_s": 0.28000000000000114, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/89f2eb321df0d91bf2d31124360b336d1ac4f912dfad856ccfa3db142c273681.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "89fcdf9d000ed85a78cf32bb34dc8b3451bd1b1558d920a94560e1fec8ba2c64", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "615aa1d461aa0dfebbbb2378c5e20cb00e41842565dce7bc4121f6bf9d1d775d", "generation_seconds_batch": 24.43250141700264, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.924342393875122, "target_z": 1.6874999783840168, "ce": 6.09409761428833, "genuineness": 4.469690799713135, "identity": 0.4379187822341919, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/89fcdf9d000ed85a78cf32bb34dc8b3451bd1b1558d920a94560e1fec8ba2c64.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "89ffe9d12a1c050e25b8bc41c9a44840997a2456c8c44dfaadc49e5be58f3d62", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "42cd864330f0ce87998b2e82e2a7c1e31175a3ca279ba6b3ed2f376a05b99e01", "generation_seconds_batch": 22.534372791997157, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4180946350097656, "target_z": 1.32031238381409, "ce": 5.816103935241699, "genuineness": 4.469690799713135, "identity": 0.414808452129364, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/89ffe9d12a1c050e25b8bc41c9a44840997a2456c8c44dfaadc49e5be58f3d62.mp3" }, { "id": "heldout-nl-2", "language": "nl", "sentences": [ "De weg gaat linksaf na de brug.", "Loop verder... tot aan het kleine houten hek.", "Ik wacht aan de andere kant." ], "text": "De weg gaat linksaf na de brug. Loop verder... tot aan het kleine houten hek. Ik wacht aan de andere kant.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-2", "key": "8a0823a5054b01347fac89ee6514734da9f3baf8b31dcccb7cb20fcc2cdd832d", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik wacht aan de andere kant.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Natural clear speech.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Natural clear speech.) [3.2 seconds duration] Ik wacht aan de andere kant.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4ed822741c85ba85548749321e21713cb25db8b9913102a2ff86c76119663b10", "generation_seconds_batch": 14.974486549966969, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.019559383392334, "target_z": 1.0312499675760252, "ce": 5.824047088623047, "genuineness": 4.432034492492676, "identity": 0.3883855938911438, "hypothesis": "De weg gaat linksaf naar de brug. Loop verder tot aan het kleine houten hek. Ik wacht aan de andere kant.", "wer": 0.047619047619047616, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8a0823a5054b01347fac89ee6514734da9f3baf8b31dcccb7cb20fcc2cdd832d.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "8a0b9e8a5d4a723c08d038c28e4eebcdde16716237c30968efa6dff9b3086b0e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b52876e8e89b4c7fdcaf56a347c4aef0306aa982b018c9801f50302f3b49f65c", "generation_seconds_batch": 23.071059967041947, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8274013996124268, "target_z": 1.6171875081059937, "ce": 5.919358730316162, "genuineness": 4.65797233581543, "identity": 0.670447826385498, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8a0b9e8a5d4a723c08d038c28e4eebcdde16716237c30968efa6dff9b3086b0e.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "8a4d3bccc4875112e8b5912c5e91eb141f1aa9e904927e8cbff7b270656dfec4", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5a55590951e5fab97f35cb0db376113bed30544b8201efd061d251b3bd171f56", "generation_seconds_batch": 23.775579878943972, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.346357822418213, "target_z": 0.5429687702649842, "ce": 6.260893821716309, "genuineness": 4.406930446624756, "identity": 0.6813292503356934, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8a4d3bccc4875112e8b5912c5e91eb141f1aa9e904927e8cbff7b270656dfec4.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "8a8b960b895f74452801d76ea701f50cab0427fc76d073e6e6473a21cd22358c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "00522038cc56b48305f327515b72255f5fe8b49fb9437bd1e5fdec2686ba42df", "generation_seconds_batch": 25.4117253680015, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.169601321220398, "target_z": -0.31054688783449, "ce": 6.157639026641846, "genuineness": 4.444586753845215, "identity": 0.6819384694099426, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8a8b960b895f74452801d76ea701f50cab0427fc76d073e6e6473a21cd22358c.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "8aa83fd3515f88aa69839d05416f123c3ccd7416606596b723f0d6230a0ed09c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "a8491a9606a9bf1e6505ad70880fb262a11b7d2e7f35e95053127f008e8369ca", "generation_seconds_batch": 24.09285667806398, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.773545265197754, "target_z": 1.578125005403996, "ce": 6.022613525390625, "genuineness": 4.281409740447998, "identity": 0.5900073647499084, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8aa83fd3515f88aa69839d05416f123c3ccd7416606596b723f0d6230a0ed09c.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "8ac220b67dabfecc27f8d1321422e3a42a25bf90aece0f43c3323ecf3cd1f170", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "eab80b3f92568fa6f8b8025a1357ff9d466bfd43d32ac796d9041e017d5ca2d7", "generation_seconds_batch": 22.12326556397602, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.482722282409668, "target_z": 1.3671875945699268, "ce": 6.24500846862793, "genuineness": 4.7332844734191895, "identity": 0.5924438238143921, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8ac220b67dabfecc27f8d1321422e3a42a25bf90aece0f43c3323ecf3cd1f170.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "8b0a028abf4d069e8c8f3b583d839aebfeca852b0f4edf6fcd5f27da7bd5d878", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "99b0910c3459a1c0e00b6040e0a1fb41d71add4d207d6781c02415b9ec3b2051", "generation_seconds_batch": 24.176370520028286, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.432527542114258, "target_z": 0.6054687054170346, "ce": 6.229123115539551, "genuineness": 4.519899368286133, "identity": 0.706287145614624, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8b0a028abf4d069e8c8f3b583d839aebfeca852b0f4edf6fcd5f27da7bd5d878.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "8b14f7cc1058fed326e5c0de398b94503a5187c477f9c50d78756921ad0f36d6", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f223484b1d9e39caf3a16b6c93c24d2aa84a71d06609320cef26bd697fa5c0a0", "generation_seconds_batch": 24.186401372076944, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9172327518463135, "target_z": 0.9570312297350158, "ce": 6.1020402908325195, "genuineness": 4.595211505889893, "identity": 0.7474648952484131, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8b14f7cc1058fed326e5c0de398b94503a5187c477f9c50d78756921ad0f36d6.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "8b376eca7a1066d1c70e893f80cff7eba545e1a0f912c5f1debdbcc074163f23", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "954a39e2168d53af6461dc2e158cc3d4847e28e62d91844eb2785c590dc11ced", "generation_seconds_batch": 23.613548352965154, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.139766693115234, "target_z": 1.8437498162641426, "ce": 5.383228778839111, "genuineness": 4.419482707977295, "identity": 0.4481366276741028, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que ha.", "wer": 0.11764705882352941, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8b376eca7a1066d1c70e893f80cff7eba545e1a0f912c5f1debdbcc074163f23.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "8bb930591cccde56c29c964c33d1856e00e124c1c19e8131dad60b237fdfcd8c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "bc27a165596ccbe7ee80f6c532e2b6bc314e311fd34b0739e88dc79d3ad9a5b9", "generation_seconds_batch": 24.505743153975345, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.527531147003174, "target_z": 2.1250000432319665, "ce": 5.450741291046143, "genuineness": 4.3881025314331055, "identity": 0.84657883644104, "hypothesis": "La porte est encore ouverte, j'ai posé ton manteau sur la chaise, dis-moi quand tu es prête.", "wer": 0.05555555555555555, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8bb930591cccde56c29c964c33d1856e00e124c1c19e8131dad60b237fdfcd8c.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "8bd9973168253f9145520c245afe26c4ee507e9d2f69874a1cce1719f1f3c68d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0fd5cc580179a0b55f8cdc4bba6448c99289793c84b570a9973a70aca4b5171c", "generation_seconds_batch": 17.477941902936436, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.364238739013672, "target_z": 1.281250054039958, "ce": 6.141753673553467, "genuineness": 4.7081804275512695, "identity": 0.6894887089729309, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8bd9973168253f9145520c245afe26c4ee507e9d2f69874a1cce1719f1f3c68d.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "8be8f80bb314cc74490a253bb0ea0d297da89d50a9d8b83f9c215dda53b94fdc", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7daf81fd1c415fdb410972155f052da4578f9f83f4df088cb0c4cfd3fe0f7d8f", "generation_seconds_batch": 17.416993562947027, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3211536407470703, "target_z": 1.249999913536067, "ce": 5.704906940460205, "genuineness": 4.080576419830322, "identity": 0.4270852208137512, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8be8f80bb314cc74490a253bb0ea0d297da89d50a9d8b83f9c215dda53b94fdc.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "8caa7753597bd9010a3e1af3eff1bf70dd168d70121f77b38ce2ac68e69dc924", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "30c9240c6a33188d4187411f698216d9db7a97095585b09f0471787909e96f57", "generation_seconds_batch": 23.542010444914922, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.470226764678955, "target_z": 0.6328124054300733, "ce": 6.078212261199951, "genuineness": 4.413206577301025, "identity": 0.6909965872764587, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8caa7753597bd9010a3e1af3eff1bf70dd168d70121f77b38ce2ac68e69dc924.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "8ce7b7082a698d3bd72b6a3b176584256bcf480682652dda7c1ee6aed60a1778", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "39698564db2ad2880b5bb7428e8aabab3b1616a1ca6809f09448eda2630bd1f1", "generation_seconds_batch": 23.77853731403593, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4934933185577393, "target_z": 1.3749999567680335, "ce": 5.974957466125488, "genuineness": 4.369274139404297, "identity": 0.4278554916381836, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8ce7b7082a698d3bd72b6a3b176584256bcf480682652dda7c1ee6aed60a1778.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "8d314515610fae1d45ff67ab2bd4f200b29bd52cec6a12a28cc3d3d2f06d564f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1c9c7292a62f2f5f25c89d8a377b385f9d04fcaebeb56adbf1340d474624fc5c", "generation_seconds_batch": 24.00689779908862, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.459455728530884, "target_z": 0.6250000432319665, "ce": 6.332377910614014, "genuineness": 4.532451152801514, "identity": 0.29273492097854614, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8d314515610fae1d45ff67ab2bd4f200b29bd52cec6a12a28cc3d3d2f06d564f.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "8da8323e27e4e156dcc13db718c65dc919a3e20bc798809452ade2e1f85ab9df", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "fae2de0118eb6be64b63bea988bec8abdff2c1e9cf3cf8af687b360ee08042c5", "generation_seconds_batch": 23.290174574009143, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1363189220428467, "target_z": 0.39062502701997903, "ce": 6.268836498260498, "genuineness": 4.519899368286133, "identity": 0.7464474439620972, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8da8323e27e4e156dcc13db718c65dc919a3e20bc798809452ade2e1f85ab9df.mp3" }, { "id": "heldout-en-2", "language": "en", "sentences": [ "The road turns left behind the bridge.", "Keep going... until you reach the small wooden gate.", "I will meet you on the other side." ], "text": "The road turns left behind the bridge. Keep going... until you reach the small wooden gate. I will meet you on the other side.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-2", "key": "8db84d8bfdd675d5261674265afb1e873893d44a1ca3775db0dbbd50350b4a3b", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The road turns left behind the bridge.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I will meet you on the other side.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The road turns left behind the bridge.\n(Natural clear speech.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Natural clear speech.) [3.2 seconds duration] I will meet you on the other side.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "96180878fc30cc810dffd95f22e8fee4decf867b03266e86669d2f72a677276c", "generation_seconds_batch": 17.88071251299698, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.342696189880371, "target_z": 1.2656249837880125, "ce": 6.125868320465088, "genuineness": 4.369274139404297, "identity": 0.3965182900428772, "hypothesis": "The road turns left behind the bridge. Keep going until you reach the small wooden gate. I will meet you on the other side.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8db84d8bfdd675d5261674265afb1e873893d44a1ca3775db0dbbd50350b4a3b.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "8df746b0f597a592880dde43b3eb9444f70a69608ea6b183e4ca5d249edb7aca", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f5d5073e2424e238d88a50e9bc6db004e9acceed43412c078ffcbc6ccbb5012e", "generation_seconds_batch": 17.126916994107887, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7304604053497314, "target_z": 1.5468750378279708, "ce": 6.308549880981445, "genuineness": 4.7834930419921875, "identity": 0.7554929256439209, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8df746b0f597a592880dde43b3eb9444f70a69608ea6b183e4ca5d249edb7aca.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "8e21a1f90e2876bbabddb073ceeb6873b74a169b82e095de07409b594121dee8", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "71cf041c22321657258cc441737f1a8e16a9b8d5e8ccbf60084e960d153a621a", "generation_seconds_batch": 22.32682138902601, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.107453346252441, "target_z": 1.820312556741956, "ce": 5.629451274871826, "genuineness": 4.632867813110352, "identity": 0.47044631838798523, "hypothesis": "He encontrado la cajita, estaba junto a la ventana vieja, podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8e21a1f90e2876bbabddb073ceeb6873b74a169b82e095de07409b594121dee8.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "8e3b2a037a4051066b70fe585b6a51b84e9920debad3d39039145d9ad9e06ef9", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "80934cd9ed3e7d6ef0c7bf3bb21d2cc6e3426e7350d1e28a1037839a8126cd03", "generation_seconds_batch": 23.431125467061065, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.269021987915039, "target_z": 1.9375002377758157, "ce": 5.808161735534668, "genuineness": 4.457139015197754, "identity": 0.22595131397247314, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8e3b2a037a4051066b70fe585b6a51b84e9920debad3d39039145d9ad9e06ef9.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "8e72c105bec268f6a1f32005a77a40bf9f5f8c64abf67ed022fcd3cec505c7f2", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "03d95513cf9d49f73e43735000ab38f24034d8dfd03eb6b1c95edfdb6e31f747", "generation_seconds_batch": 15.358593748998828, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.859714984893799, "target_z": 1.640624940556046, "ce": 5.685050010681152, "genuineness": 4.444586753845215, "identity": 0.33525794744491577, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8e72c105bec268f6a1f32005a77a40bf9f5f8c64abf67ed022fcd3cec505c7f2.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "8ec7cea853b68600d6f906d24cf8d05bda3d6aff9dc4697df6d1602ea1338288", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "a8a70478fa253e49118acccad57f4bf1d7f778b35f6c30c521fad84406616bf3", "generation_seconds_batch": 17.2146433070302, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.150538444519043, "target_z": 1.851562697245847, "ce": 6.141753673553467, "genuineness": 4.557555198669434, "identity": 0.6626026034355164, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8ec7cea853b68600d6f906d24cf8d05bda3d6aff9dc4697df6d1602ea1338288.mp3" }, { "id": "heldout-es-1", "language": "es", "sentences": [ "Las llaves están en la mesa de la cocina.", "Miré en el cajón... y luego junto a la ventana.", "Llévatelas cuando te vayas." ], "text": "Las llaves están en la mesa de la cocina. Miré en el cajón... y luego junto a la ventana. Llévatelas cuando te vayas.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-1", "key": "8eead2da5ca94c540bc5d16ccd5223dd03f1840aad69916b6f2be7a0a80f78b8", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Llévatelas cuando te vayas.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Natural clear speech.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Natural clear speech.) [3.2 seconds duration] Llévatelas cuando te vayas.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "fa7a5b95415ca249154fbec2a1dd4138086883b31ec8cf47f66f356f02c383ab", "generation_seconds_batch": 14.916390712023713, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.633519411087036, "target_z": 1.4765625675499476, "ce": 5.510311126708984, "genuineness": 4.356721878051758, "identity": 0.4928423762321472, "hypothesis": "Las llaves están en la mesa de la cocina, mire en el cajón y luego, junto a la ventana, llévatelas cuando te vayas.", "wer": 0.043478260869565216, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8eead2da5ca94c540bc5d16ccd5223dd03f1840aad69916b6f2be7a0a80f78b8.mp3" }, { "id": "heldout-en-3", "language": "en", "sentences": [ "Someone moved the little blue chair.", "It used to stand here... beside the tall bookshelf.", "We can put it back after lunch." ], "text": "Someone moved the little blue chair. It used to stand here... beside the tall bookshelf. We can put it back after lunch.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-3", "key": "8f96d177c7696c8c12783023c194864c129d325a3bfe8d7af144fa07ecc5090b", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Someone moved the little blue chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can put it back after lunch.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Someone moved the little blue chair.\n(Natural clear speech.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Natural clear speech.) [3.2 seconds duration] We can put it back after lunch.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "1ae8ab92e14ad923b718b5e28d4bfd8d8e57177618c388b5bf18b8af574254f7", "generation_seconds_batch": 16.48735069297254, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5796632766723633, "target_z": 1.4375000648479497, "ce": 5.871703147888184, "genuineness": 4.243753433227539, "identity": 0.34499257802963257, "hypothesis": "Someone moved the little blue chair. It used to stand here beside the tall bookshelf. We can put it back after lunch.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8f96d177c7696c8c12783023c194864c129d325a3bfe8d7af144fa07ecc5090b.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "8faa6ca2b9b63e6b6112b94809541d3b4b807fa7af25614bdba7af65bbe84885", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a0dcb9a68f5e2991259300792daf1a29beebb848c42569dd436af390ed840bfb", "generation_seconds_batch": 23.951555098989047, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.450408458709717, "target_z": 1.3437499891920084, "ce": 6.300607204437256, "genuineness": 4.381826400756836, "identity": 0.2795804440975189, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8faa6ca2b9b63e6b6112b94809541d3b4b807fa7af25614bdba7af65bbe84885.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "8ff0db2f73efa313077388b2a200c28181de073d206572fc0e115d32a23a7dd6", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b34b3bd5d2a1a0c466ef0723f767491389343d474ba976cc2bb603dab25d0f8a", "generation_seconds_batch": 24.03318282496184, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3140439987182617, "target_z": 0.519531164887066, "ce": 6.0146708488464355, "genuineness": 4.6705241203308105, "identity": 0.5159590840339661, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/8ff0db2f73efa313077388b2a200c28181de073d206572fc0e115d32a23a7dd6.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "910077c9d32fa74dca1af6238d55c78f55e9574536f36e3dea7e3270463aac8d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c998be6195e5c0e4064895bfcca61be71c0014a3acef8c918d491b612ab0a1bc", "generation_seconds_batch": 25.48447667597793, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2349839210510254, "target_z": 1.1874999783840168, "ce": 5.879645347595215, "genuineness": 4.519899368286133, "identity": 0.7099797129631042, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás lista.", "wer": 0.06666666666666667, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/910077c9d32fa74dca1af6238d55c78f55e9574536f36e3dea7e3270463aac8d.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "917a51fbb057a3cd25fdad21ee504c29bc3e6f4bc6d2dfd4e9d307f885ca62f4", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "89d5e90cded72c5f54c665b21617eb8de1142b527d344aeac9654ec597161a62", "generation_seconds_batch": 21.842607220984064, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.123823642730713, "target_z": -0.3437499891920084, "ce": 6.125868320465088, "genuineness": 4.457139015197754, "identity": 0.7876261472702026, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/917a51fbb057a3cd25fdad21ee504c29bc3e6f4bc6d2dfd4e9d307f885ca62f4.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "918e9c4d057fef9de0c32cc473de21eb682708cf02f3756e463472ee3e2444d2", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3d1640b6c9c1cd2a902be8c228df184385947f440f6e3abf2005a9a8c5b61275", "generation_seconds_batch": 22.669165586004965, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7071938514709473, "target_z": 0.8046874864900104, "ce": 6.221180438995361, "genuineness": 4.369274139404297, "identity": 0.8553031086921692, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/918e9c4d057fef9de0c32cc473de21eb682708cf02f3756e463472ee3e2444d2.mp3" }, { "id": "heldout-de-4", "language": "de", "sentences": [ "Ich schrieb die Nummer auf den Umschlag.", "Lies sie genau... und prüfe die letzten beiden Ziffern.", "Ruf an, wenn du Zeit hast." ], "text": "Ich schrieb die Nummer auf den Umschlag. Lies sie genau... und prüfe die letzten beiden Ziffern. Ruf an, wenn du Zeit hast.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-4", "key": "935c3003626cc102d4b3fde8f8161ba8e7550aef58f303f9367dc8adfefe14f1", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Natural clear speech.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Natural clear speech.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "60812fbd817de6d78cec684c54a3ae821d8406a7521803e8e1398141abb29e9a", "generation_seconds_batch": 23.28906766604632, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.547349452972412, "target_z": 1.4140624594700315, "ce": 5.974957466125488, "genuineness": 4.632867813110352, "identity": 0.6067134737968445, "hypothesis": "Ich schrieb die Nummer auf den Umschlag, lies sie genau und prüfe die letzten beiden Ziffern. Ruf an, wenn du Zeit hast.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/935c3003626cc102d4b3fde8f8161ba8e7550aef58f303f9367dc8adfefe14f1.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "93d6b9ee2817e554f1f8ec76854fa07672098f36605442842766089e8edd6062", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "07383f20cd6e2a8a7b26a50e2e0c9cb66f81a23b550e9b87ba63fd532f575703", "generation_seconds_batch": 23.892786954995245, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8212605714797974, "target_z": 0.16210941215247118, "ce": 6.395919322967529, "genuineness": 4.193545341491699, "identity": 0.09387698024511337, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/93d6b9ee2817e554f1f8ec76854fa07672098f36605442842766089e8edd6062.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "954fcacd5327f4e1f00e0d6efec629776a73bf592a8a39a55c1f4cfc75677da0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "6768de7da77f454e87f3c011a8930c41c57fdefe1e14c55cb66d8978e382007e", "generation_seconds_batch": 24.523894718964584, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.924342393875122, "target_z": 1.6874999783840168, "ce": 5.250188827514648, "genuineness": 4.419482707977295, "identity": 0.7041580677032471, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle conduit.", "wer": 0.05263157894736842, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/954fcacd5327f4e1f00e0d6efec629776a73bf592a8a39a55c1f4cfc75677da0.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "9647c1701ba9f58a660f7cf8a161b5e906c6fb01b03afa8d73ec97d9d0f489fc", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9b94ef4c333a71e1f5cfaf21c87d86d5ead72e04cd1c2dc481fc7a2befefc7be", "generation_seconds_batch": 24.81006707798224, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.945884943008423, "target_z": 1.7031250486359624, "ce": 5.967014789581299, "genuineness": 4.532451152801514, "identity": 0.5237475037574768, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9647c1701ba9f58a660f7cf8a161b5e906c6fb01b03afa8d73ec97d9d0f489fc.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "9650aca6ee28a530fadce8e134524c3f10306473d17254ec373081e34754d697", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1a071c9e9b9c7985d494b1b5e04de5d2639f65082add6c63dc07a14bed56923a", "generation_seconds_batch": 16.066815444966778, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.859714984893799, "target_z": 1.640624940556046, "ce": 5.839932441711426, "genuineness": 4.494794845581055, "identity": 0.7668520212173462, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9650aca6ee28a530fadce8e134524c3f10306473d17254ec373081e34754d697.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "971e205b375f3390a49196fcbba7d9f365acc64bc8c1d3780e9459d7dfa79ac2", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "56eaf1ebe7c1e48032c1c6909830ea0b4894a209e9f24bc1237a48324f87ebb1", "generation_seconds_batch": 22.88479917193763, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.139766693115234, "target_z": 1.8437498162641426, "ce": 6.0146708488464355, "genuineness": 4.607763767242432, "identity": 0.7441684007644653, "hypothesis": "La porte est encore ouverte, j'ai posé ton manteau sur la chaise, dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/971e205b375f3390a49196fcbba7d9f365acc64bc8c1d3780e9459d7dfa79ac2.mp3" }, { "id": "heldout-en-2", "language": "en", "sentences": [ "The road turns left behind the bridge.", "Keep going... until you reach the small wooden gate.", "I will meet you on the other side." ], "text": "The road turns left behind the bridge. Keep going... until you reach the small wooden gate. I will meet you on the other side.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-2", "key": "9732b3c6e2f128cb6fdbb42a05bf9f44c635b432485bf735fb9a73d78e0d83bb", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The road turns left behind the bridge.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I will meet you on the other side.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The road turns left behind the bridge.\n(Natural clear speech.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Natural clear speech.) [3.2 seconds duration] I will meet you on the other side.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f25f3141abf12abdb5ef8af6b60408212c109e581b7387e41b0b045d4e631beb", "generation_seconds_batch": 15.980758043006063, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.342696189880371, "target_z": 1.2656249837880125, "ce": 6.022613525390625, "genuineness": 4.419482707977295, "identity": 0.44416022300720215, "hypothesis": "The road turns left behind the bridge. Keep going until you reach the small wooden gate. I will meet you on the other side.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9732b3c6e2f128cb6fdbb42a05bf9f44c635b432485bf735fb9a73d78e0d83bb.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "9758298f3b570fe69f99a12357bb6f6043df4a628e56d87c17986ba60d2d6176", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "564fba59e04e9c53091bd1cf557c589ccc07cd49968c05eef32eea76249db4e7", "generation_seconds_batch": 22.486650948994793, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.342696189880371, "target_z": 1.2656249837880125, "ce": 5.5539960861206055, "genuineness": 4.413206577301025, "identity": 0.34047985076904297, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9758298f3b570fe69f99a12357bb6f6043df4a628e56d87c17986ba60d2d6176.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "9786fae314cb3fa226329f26cdeb3213e1a251851930daa117ed74487fe8b477", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "788881d9242e052ce4650cb00977fb31afcc7c665045cbcc2c6807638db23c2b", "generation_seconds_batch": 24.547520775930025, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2349839210510254, "target_z": 1.1874999783840168, "ce": 6.157639026641846, "genuineness": 4.337893962860107, "identity": 0.7974326610565186, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9786fae314cb3fa226329f26cdeb3213e1a251851930daa117ed74487fe8b477.mp3" }, { "id": "heldout-en-4", "language": "en", "sentences": [ "I wrote the number on the envelope.", "Read it carefully... and check the last two digits.", "Call me when you have a moment." ], "text": "I wrote the number on the envelope. Read it carefully... and check the last two digits. Call me when you have a moment.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-4", "key": "97c4e1c457068c225e01dd571c28929f99cefcaca1d3d59b2e1edf61eda929d8", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I wrote the number on the envelope.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Call me when you have a moment.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I wrote the number on the envelope.\n(Natural clear speech.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Natural clear speech.) [3.2 seconds duration] Call me when you have a moment.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f33c8c7f6c630b40b5595e5e889bee385b046a6be7de884ecb7c7d8376b9e77c", "generation_seconds_batch": 22.129737686016597, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.793363571166992, "target_z": 0.8671874216420608, "ce": 5.95112943649292, "genuineness": 4.111956596374512, "identity": 0.3850700557231903, "hypothesis": "I wrote the number on the envelope. Read it carefully and check the last two digits. Call me when you have a moment.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/97c4e1c457068c225e01dd571c28929f99cefcaca1d3d59b2e1edf61eda929d8.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "97ee2949e92b934fc04b4dcc5ac364150fb118b0d1ef8e42da38023571e3acf2", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9544689942200bf1a9dd6cd234dcf8d1131bdadf187febb4782a1d7bdf15b733", "generation_seconds_batch": 16.55857592704706, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.838172674179077, "target_z": 1.6250000432319665, "ce": 6.070269584655762, "genuineness": 4.482243061065674, "identity": 0.2869957387447357, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/97ee2949e92b934fc04b4dcc5ac364150fb118b0d1ef8e42da38023571e3acf2.mp3" }, { "id": "heldout-es-3", "language": "es", "sentences": [ "Alguien ha movido la pequeña silla azul.", "Estaba aquí... junto a la estantería alta.", "La colocaremos después de comer." ], "text": "Alguien ha movido la pequeña silla azul. Estaba aquí... junto a la estantería alta. La colocaremos después de comer.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-3", "key": "97efeed0223e6db2fde0e825c5329e848ce737fccee710925c92fa2592300d39", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La colocaremos después de comer.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Natural clear speech.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Natural clear speech.) [3.2 seconds duration] La colocaremos después de comer.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "068a8772f2749959b0ee6b830cc89e9f3c264dee6bffaf04bcfd9c2a1e4128d6", "generation_seconds_batch": 14.859678464010358, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6766042709350586, "target_z": 1.5078125351259728, "ce": 5.780362129211426, "genuineness": 4.7081804275512695, "identity": 0.6746172308921814, "hypothesis": "Alguien ha movido la pequeña silla azul. Estaba aquí, junto a la estantería alta. La colocaremos después de comer.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/97efeed0223e6db2fde0e825c5329e848ce737fccee710925c92fa2592300d39.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "980c63c760ee9e889c15d7f22aadf6235e57c736bb74f1df59922cec6aae565b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1ab727e9e0c805a80f40f355d6f2e70819d97c933e2b26ec05ef7609729cd012", "generation_seconds_batch": 23.671042102971114, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0411019325256348, "target_z": 1.0468750378279708, "ce": 6.2767791748046875, "genuineness": 4.432034492492676, "identity": 0.7708631157875061, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/980c63c760ee9e889c15d7f22aadf6235e57c736bb74f1df59922cec6aae565b.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "9868b99d73a2173d33bd4402b51cff9384d628c841fea06c97b6861fd1eb8397", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "bf7c7b63a6a23b32bd766284aab179154858fd7a97824dd9208e623ace1863d4", "generation_seconds_batch": 23.259458995074965, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.547349452972412, "target_z": 1.4140624594700315, "ce": 5.780362129211426, "genuineness": 4.149612903594971, "identity": 0.3556774854660034, "hypothesis": "I found the little box. It was beside the old window. We can look inside today.", "wer": 0.0625, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9868b99d73a2173d33bd4402b51cff9384d628c841fea06c97b6861fd1eb8397.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "987223aa9395d2351689c788649e7b9afc78bca2cc395601d4a50168991b0af6", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b3382d62d65f90a2543bafbfacd1c869ad87ccbc7f4e3863fcacb6a8539f6732", "generation_seconds_batch": 24.888466550968587, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7089178562164307, "target_z": 1.5312499675760252, "ce": 6.149696350097656, "genuineness": 4.582659721374512, "identity": 0.8139077425003052, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/987223aa9395d2351689c788649e7b9afc78bca2cc395601d4a50168991b0af6.mp3" }, { "id": "heldout-nl-2", "language": "nl", "sentences": [ "De weg gaat linksaf na de brug.", "Loop verder... tot aan het kleine houten hek.", "Ik wacht aan de andere kant." ], "text": "De weg gaat linksaf na de brug. Loop verder... tot aan het kleine houten hek. Ik wacht aan de andere kant.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-2", "key": "98a766ba71210f6182fe683de9385840ad54be774709500fd31877176a162662", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik wacht aan de andere kant.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Natural clear speech.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Natural clear speech.) [3.2 seconds duration] Ik wacht aan de andere kant.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5f6860d206ed343f46b077efb2a8d1f9aead97ec36ca37fa940258bd1e3b25a8", "generation_seconds_batch": 24.60657593992073, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5258069038391113, "target_z": 1.398437389218086, "ce": 5.847874641418457, "genuineness": 4.394378662109375, "identity": 0.2448495328426361, "hypothesis": "De weg gaat linksaf naar de brug. Loop verder tot aan het kleine houten hek. Ik wacht aan de andere kant.", "wer": 0.047619047619047616, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/98a766ba71210f6182fe683de9385840ad54be774709500fd31877176a162662.mp3" }, { "id": "heldout-de-3", "language": "de", "sentences": [ "Jemand hat den kleinen blauen Stuhl umgestellt.", "Er stand hier... neben dem hohen Bücherregal.", "Wir stellen ihn nach dem Essen zurück." ], "text": "Jemand hat den kleinen blauen Stuhl umgestellt. Er stand hier... neben dem hohen Bücherregal. Wir stellen ihn nach dem Essen zurück.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-3", "key": "98ca4946ac892dba890ce17ab3f9887db850ca084af2bb7f604223fb807fcaef", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Natural clear speech.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Natural clear speech.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e1bad675afe5b80ca428700fe2fb1c334ffdca6e3409ea38d837e961f5c1f80c", "generation_seconds_batch": 17.473335723974742, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0841867923736572, "target_z": 1.078125005403996, "ce": 5.2640886306762695, "genuineness": 4.256305694580078, "identity": 0.48502710461616516, "hypothesis": "Ihr Mann hat den kleinen blauen Stuhl umgestellt. Er stand hier neben dem hohen Bücherregal. Wir stellen ihn nach dem Essen zurück.", "wer": 0.09523809523809523, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/98ca4946ac892dba890ce17ab3f9887db850ca084af2bb7f604223fb807fcaef.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "9939bede7f401d4cdaf4566692de11e9721139c9747e369d573aae7e235c28fb", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5d9fa0fc87ff8c061d738c142d23808b1d662abc867d134b2ce58b7277d0bef3", "generation_seconds_batch": 24.146093792049214, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6210241317749023, "target_z": 0.7421875513379602, "ce": 6.197352409362793, "genuineness": 4.457139015197754, "identity": 0.6756348609924316, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9939bede7f401d4cdaf4566692de11e9721139c9747e369d573aae7e235c28fb.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "998aef213f37bdb429e432ab7e988e8e932f70982dc2019ef7075cef20aa33bf", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d0dd411f1ac7dadde003abb2f4937205b97dff0137cdd4b83711963ef2849cca", "generation_seconds_batch": 15.981465880991891, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.773545265197754, "target_z": 1.578125005403996, "ce": 5.871703147888184, "genuineness": 4.400654315948486, "identity": 0.5914821624755859, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/998aef213f37bdb429e432ab7e988e8e932f70982dc2019ef7075cef20aa33bf.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "999daa9203c1696fc331943083144154a32dd66e5dd8c49876c4cb5dc2d99d69", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "af3aa2eb6c0e89a1df8a40c1518b3a61386d08a04d90ad715885af85544140cd", "generation_seconds_batch": 25.056232245988213, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9674272537231445, "target_z": 1.718749945960042, "ce": 5.760505676269531, "genuineness": 4.494794845581055, "identity": 0.8238315582275391, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/999daa9203c1696fc331943083144154a32dd66e5dd8c49876c4cb5dc2d99d69.mp3" }, { "id": "heldout-de-0", "language": "de", "sentences": [ "Deine Nachricht kam, als ich schon weg war.", "Ich las sie zweimal... dann legte ich das Handy weg.", "Sag mir, wann wir reden können." ], "text": "Deine Nachricht kam, als ich schon weg war. Ich las sie zweimal... dann legte ich das Handy weg. Sag mir, wann wir reden können.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-0", "key": "99bf1b7d3eed7b885fdae888ab00b4960791c3971d259b46f771306d2801e39b", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wann wir reden können.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wann wir reden können.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f6dc552c745f01eb53a3a7aa0761a9a74b260438e2d310c26356ab052d2534fa", "generation_seconds_batch": 18.830065532005392, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8454958200454712, "target_z": 0.17968752972197696, "ce": 6.475345611572266, "genuineness": 4.7332844734191895, "identity": 0.5347562432289124, "hypothesis": "Deine Nachricht kam, als ich schon weg war. Ich las sie zweimal. Dann legte ich das Handy weg. Sag mir, wann wir reden können.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/99bf1b7d3eed7b885fdae888ab00b4960791c3971d259b46f771306d2801e39b.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "9a0b31ff94c4d22df4319f65cda466edbb3ad16e2961abc0a9325eddab975111", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "af1e4adac878a889e07ec1cf3283c846817a6c0dfb9b82f5c7d6eb572117919c", "generation_seconds_batch": 24.09275859897025, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.319429874420166, "target_z": 0.5234376053779183, "ce": 6.427689552307129, "genuineness": 4.432034492492676, "identity": 0.580341637134552, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9a0b31ff94c4d22df4319f65cda466edbb3ad16e2961abc0a9325eddab975111.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "9a3c66774333f1538fd44cddb9d9585213f82b19e4f8480b31c24ce33815c5d2", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4d5e5fb875b67b8d3f5704207d2e920299ae9ab0b674529b039b022d00a03f10", "generation_seconds_batch": 24.21342054498382, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0841867923736572, "target_z": 1.078125005403996, "ce": 6.173524379730225, "genuineness": 4.482243061065674, "identity": 0.4654356837272644, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9a3c66774333f1538fd44cddb9d9585213f82b19e4f8480b31c24ce33815c5d2.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "9b00feb310b463308fa42c15ab879176a115fac1d7f0fc88129a980343e2c12f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.0, "cfg": 2.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "bf3c4a8518d482f3e5cc359aed28922199c231bcd56879321a2fa3e13c61820e", "generation_seconds_batch": 18.789165863068774, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.105729103088379, "target_z": 1.0937499027280755, "ce": 6.09409761428833, "genuineness": 4.507347106933594, "identity": 0.8314803838729858, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9b00feb310b463308fa42c15ab879176a115fac1d7f0fc88129a980343e2c12f.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "9bdab9e3a70726b8504cb9c23527df436e05bd0e7104e70a47696138e2a49f69", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ed2f9e6ddecd08e7d3f2481021e621a0061f249007f1778b0f194f0f90d43a27", "generation_seconds_batch": 32.573457063990645, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.647952079772949, "target_z": 0.7617187162250262, "ce": 5.959072113037109, "genuineness": 4.519899368286133, "identity": 0.7179292440414429, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9bdab9e3a70726b8504cb9c23527df436e05bd0e7104e70a47696138e2a49f69.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "9bfec6f038642d0d2385cb7250b3709f479270e48a1c8c53f489fc01e5a88b30", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b55dbd482b876e6953d760d0fa7a3bab856ece9573f19d04e6d7aad13ba74627", "generation_seconds_batch": 16.22605377691798, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2349839210510254, "target_z": 1.1874999783840168, "ce": 6.006728172302246, "genuineness": 4.444586753845215, "identity": 0.5991759896278381, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9bfec6f038642d0d2385cb7250b3709f479270e48a1c8c53f489fc01e5a88b30.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "9c0847b6984efe3ddc6f1cf80b59a7a49a2c7ad95b803a8cf551ce006fc349b7", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "272d5b3f8bfb0d6f825c32601b19baddf8c7f5b042c88dd0ecb5678525fe1409", "generation_seconds_batch": 17.671789115993306, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9380199909210205, "target_z": -0.4785156148675079, "ce": 6.141753673553467, "genuineness": 4.519899368286133, "identity": 0.7431647181510925, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9c0847b6984efe3ddc6f1cf80b59a7a49a2c7ad95b803a8cf551ce006fc349b7.mp3" }, { "id": "heldout-nl-3", "language": "nl", "sentences": [ "Iemand heeft het kleine blauwe stoeltje verplaatst.", "Het stond hier... naast de hoge boekenkast.", "We zetten het na de lunch terug." ], "text": "Iemand heeft het kleine blauwe stoeltje verplaatst. Het stond hier... naast de hoge boekenkast. We zetten het na de lunch terug.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-3", "key": "9c0869bef0a211dde3611405f617519592f439613241cad97fb9970068db7fe0", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We zetten het na de lunch terug.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Natural clear speech.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Natural clear speech.) [3.2 seconds duration] We zetten het na de lunch terug.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "167544f770bf4d74691880a0a31b517cecd44bfea6f5fb909bd5cf6decb8356c", "generation_seconds_batch": 14.943401627941057, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.965703248977661, "target_z": 0.9921874648740272, "ce": 6.221180438995361, "genuineness": 4.7081804275512695, "identity": 0.44360417127609253, "hypothesis": "Iemand heeft het kleine blauwe stoeltje verplaatst. Het stond hier naast de hoge boekenkast. We zetten het na de lunch terug.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9c0869bef0a211dde3611405f617519592f439613241cad97fb9970068db7fe0.mp3" }, { "id": "heldout-nl-3", "language": "nl", "sentences": [ "Iemand heeft het kleine blauwe stoeltje verplaatst.", "Het stond hier... naast de hoge boekenkast.", "We zetten het na de lunch terug." ], "text": "Iemand heeft het kleine blauwe stoeltje verplaatst. Het stond hier... naast de hoge boekenkast. We zetten het na de lunch terug.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-3", "key": "9c09e940c5dab53287ffae2d1a8fa5296f41b2df71714656aaa2bb8830bf5162", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We zetten het na de lunch terug.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We zetten het na de lunch terug.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "30d60a366e16281d642747a8e20207d01a472d3a56f5e5508473dc75a9aa22c9", "generation_seconds_batch": 17.332933452911675, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9622552394866943, "target_z": -0.4609374972980021, "ce": 6.205295085906982, "genuineness": 4.406930446624756, "identity": 0.7395548820495605, "hypothesis": "Iemand heeft het kleine blauwe stoeltje verplaatst. Het stond hier naast de hoge boekenkast. We zetten het na de lunch terug.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9c09e940c5dab53287ffae2d1a8fa5296f41b2df71714656aaa2bb8830bf5162.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "9c9332ca66e4c95f4287d0693e4d56c56345da154032f837cc6d7762e0ba599e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0966d3b3bf8a0742c9e26751b8fd16e6823bb6b23d3593e1cfab7ae394c78811", "generation_seconds_batch": 22.61160150100477, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.570615768432617, "target_z": 2.1562498378801256, "ce": 5.720792293548584, "genuineness": 4.482243061065674, "identity": 0.6269201636314392, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9c9332ca66e4c95f4287d0693e4d56c56345da154032f837cc6d7762e0ba599e.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "9c9574ab2a657ac2cfaf4e8dd30637dbac9feaf38dd4afb07f9c6d2514e7a42b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "14965b997f8ae5cb4adad327993abca2d30756bc61481bc76e3c66771a4770b3", "generation_seconds_batch": 16.83830051100813, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8077964782714844, "target_z": 0.15234374324500524, "ce": 6.181467056274414, "genuineness": 4.281409740447998, "identity": 0.8152620196342468, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9c9574ab2a657ac2cfaf4e8dd30637dbac9feaf38dd4afb07f9c6d2514e7a42b.mp3" }, { "id": "heldout-es-2", "language": "es", "sentences": [ "La carretera gira a la izquierda tras el puente.", "Sigue... hasta la pequeña puerta de madera.", "Te esperaré al otro lado." ], "text": "La carretera gira a la izquierda tras el puente. Sigue... hasta la pequeña puerta de madera. Te esperaré al otro lado.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-2", "key": "9c969db20a6db380e08592f6aed9c9189b3db634d677ebe8f573f593ebdfd2d3", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Te esperaré al otro lado.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Te esperaré al otro lado.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "fe22d09d6e5b8e1a7148a705bbb25aea872a1e215598f9c24e330284602fdfa0", "generation_seconds_batch": 17.443841414991766, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7081626653671265, "target_z": 0.080078139185489, "ce": 6.141753673553467, "genuineness": 4.532451152801514, "identity": 0.44787296652793884, "hypothesis": "La carretera gira a la izquierda tras el puente. Sigue hasta la pequeña puerta de madera. Te esperaré al otro lado.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9c969db20a6db380e08592f6aed9c9189b3db634d677ebe8f573f593ebdfd2d3.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "9cb1cb56d72212fcb53098b33c8c35796880a16dfc8e67c2e205ddb54b12e73a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "11a58048fea3b20626abde87a85250f7c0dc0d5d1312c9aa14de32db08f57adc", "generation_seconds_batch": 24.34498886798974, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.665832996368408, "target_z": 1.5, "ce": 5.847874641418457, "genuineness": 4.457139015197754, "identity": 0.5906493067741394, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9cb1cb56d72212fcb53098b33c8c35796880a16dfc8e67c2e205ddb54b12e73a.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "9d20d9560139ab1db71cdbe61e930431a1fb9015296f17cee5e0aac5ee821411", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9dca75f53ce4131a9065523c1d4db75263b5a58ca0f3c4e24a07eda302547c86", "generation_seconds_batch": 24.84561343002133, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.691037178039551, "target_z": 0.7929688567289173, "ce": 5.617537498474121, "genuineness": 4.093128204345703, "identity": 0.5220345854759216, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás lista.", "wer": 0.06666666666666667, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9d20d9560139ab1db71cdbe61e930431a1fb9015296f17cee5e0aac5ee821411.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "9d856f6436ad5d57d7603b74dad88ad43795c36ee9972650e0b747a128d9449f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b8c99411388f3e3f1cb1f5d83e0ed8f16a110dbf08a4f37de6387b2861c20fa8", "generation_seconds_batch": 14.67994263197761, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.405599594116211, "target_z": 0.5859375405299686, "ce": 6.173524379730225, "genuineness": 4.325342178344727, "identity": 0.6708309650421143, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9d856f6436ad5d57d7603b74dad88ad43795c36ee9972650e0b747a128d9449f.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "9dab74e13fdf6bcc8a26dbce2aa3271b265b27b4959df4eba1b6593c01afc6ea", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "358130aa21628226a67ce94be7603d692a6c9113e3877cdc4399c325e01fff2f", "generation_seconds_batch": 24.829344674013555, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7089178562164307, "target_z": 1.5312499675760252, "ce": 5.331601142883301, "genuineness": 4.20609712600708, "identity": 0.576134443283081, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9dab74e13fdf6bcc8a26dbce2aa3271b265b27b4959df4eba1b6593c01afc6ea.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "9de08dd555b0bed89525ade6504bf3b086d711f379882ab7ed0e2a53fa8a0631", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "cd17b03854036e09c26d3ad984f8a4032d465bff1a3f973af47434500a63b7df", "generation_seconds_batch": 17.33628339599818, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.633519411087036, "target_z": 1.4765625675499476, "ce": 5.982900142669678, "genuineness": 4.494794845581055, "identity": 0.6876679062843323, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás lista.", "wer": 0.06666666666666667, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9de08dd555b0bed89525ade6504bf3b086d711f379882ab7ed0e2a53fa8a0631.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "9de50cbecc6cb038636fd4306355569611057ba5f8ffc7cac0adc5cda20799f3", "bucket": "vn_S_ASMR_reg", "phase": "wording", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "49e8f1ccf9cab25bb216d1ab7bcbf8097bd4063a367e08b4255dd133ec943383", "generation_seconds_batch": 15.99711354705505, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7162411212921143, "target_z": 0.08593754052996858, "ce": 5.974957466125488, "genuineness": 3.986435890197754, "identity": 0.3146428167819977, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás lista.", "wer": 0.06666666666666667, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9de50cbecc6cb038636fd4306355569611057ba5f8ffc7cac0adc5cda20799f3.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "9e4647053e9b82a9f41983a88cdaa8930a7ba999e514af6d78321f1144f74bb8", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "21d4ad5dd6d7939f9e78c4b43598bbca98320b675d4138948589cbcd4bbda533", "generation_seconds_batch": 23.977246632101014, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.549073219299316, "target_z": 2.14062476762818, "ce": 5.418970584869385, "genuineness": 4.507347106933594, "identity": 0.7311756610870361, "hypothesis": "J'ai trouvé la petite boîte. It's it pretty de la vieille fenêtre. No regard ensemble the cell contiene.", "wer": 0.47368421052631576, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9e4647053e9b82a9f41983a88cdaa8930a7ba999e514af6d78321f1144f74bb8.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "9ee86a08bcc5bb636eacf497b8c0ce16e909b810a658c2762df15a6dcd91686f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "41613c3a1f141a76f3509d7468ef55237d98fa9ef8cd825944f7f0e53a241d22", "generation_seconds_batch": 23.25922657107003, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.139766693115234, "target_z": 1.8437498162641426, "ce": 5.319687366485596, "genuineness": 4.331617832183838, "identity": 0.7334826588630676, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9ee86a08bcc5bb636eacf497b8c0ce16e909b810a658c2762df15a6dcd91686f.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "9ef88a2730f7012c9a91316d0a7ac8fcea2e8a34beb6ad3fc41463a98315461c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "81ec6f1d88e6079c53e419bc2c5b7e86063850c6ec9ff936b7c55fc55b4c55cb", "generation_seconds_batch": 23.054025576915592, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.419818878173828, "target_z": 2.0468750378279705, "ce": 5.808161735534668, "genuineness": 4.507347106933594, "identity": 0.628970742225647, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9ef88a2730f7012c9a91316d0a7ac8fcea2e8a34beb6ad3fc41463a98315461c.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "9efff57ac9b0535f813d46a1f966c97e9cb283e6eda32d992729a6e6ea79c0d0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "826482b4a8606d0b24937a34775a6e42e255512ad85014e9a0de254655be0fc9", "generation_seconds_batch": 17.16528386902064, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4396371841430664, "target_z": 1.3359374540660356, "ce": 6.181467056274414, "genuineness": 4.281409740447998, "identity": 0.2595025599002838, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9efff57ac9b0535f813d46a1f966c97e9cb283e6eda32d992729a6e6ea79c0d0.mp3" }, { "id": "heldout-es-2", "language": "es", "sentences": [ "La carretera gira a la izquierda tras el puente.", "Sigue... hasta la pequeña puerta de madera.", "Te esperaré al otro lado." ], "text": "La carretera gira a la izquierda tras el puente. Sigue... hasta la pequeña puerta de madera. Te esperaré al otro lado.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-2", "key": "9f80ce0459e197e3e309d556d4359ce216712d81cc844b3d7055351d4ea43898", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Te esperaré al otro lado.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Natural clear speech.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Natural clear speech.) [3.2 seconds duration] Te esperaré al otro lado.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "946c621edb151f1dd1081987899ccd632133f7ced2f3be336d3eebb9e5eb69f7", "generation_seconds_batch": 14.928427975974046, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.364238739013672, "target_z": 1.281250054039958, "ce": 5.569881439208984, "genuineness": 4.406930446624756, "identity": 0.36564603447914124, "hypothesis": "La carretera gira a la izquierda tras el puente. Sigue hasta la pequeña puerta de madera. Te esperaré al otro lado.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9f80ce0459e197e3e309d556d4359ce216712d81cc844b3d7055351d4ea43898.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "9f9397cf4b8cfab8ce325e5feb4aab5e73e2dc063a375de3ae15b9c7ee45fd00", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "d2f626b56aefb0cd738d2a5fac1045a609506d83ef7647da2a03fa9fd16d0d17", "generation_seconds_batch": 19.295539007987827, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5581207275390625, "target_z": 1.421874994596004, "ce": 5.470598220825195, "genuineness": 4.256305694580078, "identity": 0.6506391763687134, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9f9397cf4b8cfab8ce325e5feb4aab5e73e2dc063a375de3ae15b9c7ee45fd00.mp3" }, { "id": "heldout-fr-3", "language": "fr", "sentences": [ "Quelqu’un a déplacé la petite chaise bleue.", "Elle était ici... près de la grande bibliothèque.", "Nous la remettrons après le déjeuner." ], "text": "Quelqu’un a déplacé la petite chaise bleue. Elle était ici... près de la grande bibliothèque. Nous la remettrons après le déjeuner.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-3", "key": "9f95966fb1c4543256a5c3bfbc85cfb91c66b7526f44d76562b34432e54c0d44", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Natural clear speech.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Natural clear speech.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "92493e0a081590f76229e59ba5b6fc103b465882b6082229e34b9772c414e16d", "generation_seconds_batch": 15.300432551070116, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8812575340270996, "target_z": 1.6562500108079916, "ce": 5.760505676269531, "genuineness": 4.507347106933594, "identity": 0.8318530917167664, "hypothesis": "Quelqu'un a déplacé la petite chaise bleue. Elle était ici, près de la grande bibliothèque. Nous la remettrons après le déjeuner.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9f95966fb1c4543256a5c3bfbc85cfb91c66b7526f44d76562b34432e54c0d44.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "9fd2e245938e4e39c478e239ffcb4c364612e40155c0f5623eacfd64b92afe26", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "d78a251b5435ed537af24717a35ab361619b550fbe50fd240e442a14fab74909", "generation_seconds_batch": 23.870138016995043, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2349839210510254, "target_z": 1.1874999783840168, "ce": 5.033751010894775, "genuineness": 3.9173994064331055, "identity": 0.30737486481666565, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9fd2e245938e4e39c478e239ffcb4c364612e40155c0f5623eacfd64b92afe26.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "9fd74559f7d29f490054b435bd0539cb413e74f4643ed168dfb4d55f48af30a0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8fc23c730d67631c8057204388fb89071f19603021087cafcc1b9c52bcbf2e50", "generation_seconds_batch": 16.851059868000448, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.019559383392334, "target_z": 1.0312499675760252, "ce": 6.0305562019348145, "genuineness": 4.394378662109375, "identity": 0.6410475969314575, "hypothesis": "Die Tür ist noch offen. Ich hab deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.05263157894736842, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9fd74559f7d29f490054b435bd0539cb413e74f4643ed168dfb4d55f48af30a0.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "9feebebe26ac0831c90e8413d410050baa32dec28f38bb7d2fc26189f81ace7f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7f818bce601f6e7c021294c49a84311c26386c6edd79495f5a97b58494fd3136", "generation_seconds_batch": 16.75300578493625, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2996113300323486, "target_z": 1.2343750162119875, "ce": 5.174733638763428, "genuineness": 4.124508857727051, "identity": 0.3577801585197449, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás.", "wer": 0.06666666666666667, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9feebebe26ac0831c90e8413d410050baa32dec28f38bb7d2fc26189f81ace7f.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "9fef67c8721600df384ad40f6ecdcbe36b8717988f49a5a68f296a160b162fbb", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 126, "seconds_generated": 10.08, "mp3_sha256": "fd7c429441b5a127279da9d57f2d05b0d22ad82817db636354c60bf6d4f365ce", "generation_seconds_batch": 16.064213418052532, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.924342393875122, "target_z": 1.6874999783840168, "ce": 5.510311126708984, "genuineness": 4.425758361816406, "identity": 0.3718252182006836, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay.", "wer": 0.058823529411764705, "duration_error_s": 0.11999999999999922, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/9fef67c8721600df384ad40f6ecdcbe36b8717988f49a5a68f296a160b162fbb.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "a000bd552821c47511053d9ba5546d2a2433157c00b2e1ddc513ffd6ea66a178", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "ef4cdcf860cc475113597b56e9da4d596f001da3eb88fa950f755d318e0b21ee", "generation_seconds_batch": 23.114793732995167, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.687375545501709, "target_z": 1.5156250702519456, "ce": 6.070269584655762, "genuineness": 4.306513786315918, "identity": 0.42664244771003723, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a000bd552821c47511053d9ba5546d2a2433157c00b2e1ddc513ffd6ea66a178.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "a039986f1f7c56ffe2e247fb3d098ea62eb557c479bf6a601cb268612f080a1c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "55d58554be14c9e77bb3abe8936ff624f914e9b392723d02e48f6b307d0ae42b", "generation_seconds_batch": 24.725788276991807, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5365781784057617, "target_z": 1.4062499243440587, "ce": 6.078212261199951, "genuineness": 4.532451152801514, "identity": 0.3079434633255005, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a039986f1f7c56ffe2e247fb3d098ea62eb557c479bf6a601cb268612f080a1c.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "a0abc31f460977e76be01a54eba71464214a16f7f27afc8976121715dc4fefda", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "887b9a22156586029626c81ca5ca57150de5ccf17adb47aa87aff1be24ad3527", "generation_seconds_batch": 22.811438572010957, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8166301250457764, "target_z": 1.6093749729800209, "ce": 6.23706579208374, "genuineness": 4.482243061065674, "identity": 0.6495212316513062, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a0abc31f460977e76be01a54eba71464214a16f7f27afc8976121715dc4fefda.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "a0f261982d9bef2c1289a20412bd430e51749a76276966ce27d46352b238e220", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2c4097bacda3742120e5a3f35f1baa0fc87add263d0aa43990f45249bcd6d738", "generation_seconds_batch": 25.092228548950516, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9532079696655273, "target_z": 0.2578124486620398, "ce": 6.284721851348877, "genuineness": 4.325342178344727, "identity": 0.5425779819488525, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a0f261982d9bef2c1289a20412bd430e51749a76276966ce27d46352b238e220.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "a0fd0a88e4304e32e764f139297fc12fa0ab9c62d3ce3633c7a4f1e13bc532a4", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "aa4649316a70f97efbe18271a11d814de08944480ecb7ac2fc9aa7d205566781", "generation_seconds_batch": 25.65067368594464, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.701808214187622, "target_z": 0.8007812189270241, "ce": 5.982900142669678, "genuineness": 4.394378662109375, "identity": 0.23649188876152039, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a0fd0a88e4304e32e764f139297fc12fa0ab9c62d3ce3633c7a4f1e13bc532a4.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "a11756a00ed6e7fa803c2f34052a4704507894e7a2abaeeefd2dc8104f636e81", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "581546306e2f626fb567371230c849f26f7b271de5561ff797bfe8625f0f23f5", "generation_seconds_batch": 24.557565840077586, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2888400554656982, "target_z": 1.2265624810860147, "ce": 5.6334228515625, "genuineness": 4.243753433227539, "identity": 0.5588333606719971, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a11756a00ed6e7fa803c2f34052a4704507894e7a2abaeeefd2dc8104f636e81.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "a130b1c5948750aa5ab85ebad00ce3d432c6ac22a80cfb4738548e6ab1145e83", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7168f0df47025b1084c09db7471c1fbadfdc721ece366df4d9ed9015657563b2", "generation_seconds_batch": 24.43354455300141, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8454958200454712, "target_z": 0.17968752972197696, "ce": 6.395919322967529, "genuineness": 4.262581825256348, "identity": 0.711686372756958, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a130b1c5948750aa5ab85ebad00ce3d432c6ac22a80cfb4738548e6ab1145e83.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "a202caa611124daeedb73449c5ad5ac4829ba22ad891404f2baa59e6d9fb0832", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5a6604278ae726795ec43d8db5cbd21128c3cc3ad14c3ef193ea5ae568b5fc76", "generation_seconds_batch": 23.598842017934658, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2978873252868652, "target_z": 0.5078125351259728, "ce": 6.38797664642334, "genuineness": 4.337893962860107, "identity": 0.3900415897369385, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a202caa611124daeedb73449c5ad5ac4829ba22ad891404f2baa59e6d9fb0832.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "a208a47c959f93fbe18bb790354f73c08acd80da8b4a23f98894f1682870604a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "be6f392c921bfc777c0cd62e1882383406feb8f6b148c985237aef2b6ec13da3", "generation_seconds_batch": 22.783314252039418, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.588710308074951, "target_z": 0.7187499459600419, "ce": 6.046441555023193, "genuineness": 3.9927117824554443, "identity": 0.7386625409126282, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a208a47c959f93fbe18bb790354f73c08acd80da8b4a23f98894f1682870604a.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "a24d874d66263f80b107f2e645caf6f4b0ebea53c04b98ebac027f65a9126d82", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "fcbf2b7037cf8f8cc91e3fd1a3dcf8f62b9b12608f0f8961aa0c541a440b5e06", "generation_seconds_batch": 23.028504373971373, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2349839210510254, "target_z": 1.1874999783840168, "ce": 6.062326908111572, "genuineness": 4.432034492492676, "identity": 0.6160510778427124, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a24d874d66263f80b107f2e645caf6f4b0ebea53c04b98ebac027f65a9126d82.mp3" }, { "id": "heldout-en-2", "language": "en", "sentences": [ "The road turns left behind the bridge.", "Keep going... until you reach the small wooden gate.", "I will meet you on the other side." ], "text": "The road turns left behind the bridge. Keep going... until you reach the small wooden gate. I will meet you on the other side.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-2", "key": "a2707887ce1e2d94a626c4e8e9afe66c00eb0d8c7ccb40d21a3378a826fcfe47", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The road turns left behind the bridge.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I will meet you on the other side.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The road turns left behind the bridge.\n(Natural clear speech.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Natural clear speech.) [3.2 seconds duration] I will meet you on the other side.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "61298fe1cfe7d9dae786c0815fad0231dfe9691e36ac22b0aec946764ca894c3", "generation_seconds_batch": 15.980758043006063, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0303306579589844, "target_z": 1.039062502701998, "ce": 6.109982967376709, "genuineness": 4.231201171875, "identity": 0.3039788007736206, "hypothesis": "The road turns left behind the bridge. Keep going until you reach the small wooden gate. I will meet you on the other side.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a2707887ce1e2d94a626c4e8e9afe66c00eb0d8c7ccb40d21a3378a826fcfe47.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "a2c6377a56d81eb139ec6b63d377429bad8944248e356eaa1e817f40cfd3925e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2ef525a8c2b4dccecc5b80481f480543b9e12a83271ab85ac6f5139bcc0dbe56", "generation_seconds_batch": 22.24161526595708, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.633519411087036, "target_z": 1.4765625675499476, "ce": 5.903473377227783, "genuineness": 4.331617832183838, "identity": 0.3395538330078125, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a2c6377a56d81eb139ec6b63d377429bad8944248e356eaa1e817f40cfd3925e.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "a38b1ea1218ea2da4a6c8580d33a0be7e1072dc880d9b7cc0e2cdd8e2459010a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f1b4d511735e6e675f08ef9aa7b38778cfb294089026b7d71e0092bdfaa607bb", "generation_seconds_batch": 21.955919508007355, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.826646089553833, "target_z": 0.16601559325152462, "ce": 6.24500846862793, "genuineness": 4.632867813110352, "identity": 0.40400972962379456, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a38b1ea1218ea2da4a6c8580d33a0be7e1072dc880d9b7cc0e2cdd8e2459010a.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "a495b3c9afe067b161a61346e2c35c39e170094e4671d884c5afa32685a76bdc", "bucket": "vn_S_ASMR_reg", "phase": "wording", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5032016cf6bf27f943280ea949f0def1072ee933f854f4f019baafc8307a3616", "generation_seconds_batch": 17.16178419603966, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3150129318237305, "target_z": -0.20507809595352253, "ce": 6.205295085906982, "genuineness": 4.519899368286133, "identity": 0.5628232955932617, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a495b3c9afe067b161a61346e2c35c39e170094e4671d884c5afa32685a76bdc.mp3" }, { "id": "heldout-es-0", "language": "es", "sentences": [ "Tu mensaje llegó después de que me fuera.", "Lo leí dos veces... y guardé el teléfono.", "Dime cuándo podemos hablar." ], "text": "Tu mensaje llegó después de que me fuera. Lo leí dos veces... y guardé el teléfono. Dime cuándo podemos hablar.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-0", "key": "a5263b5b3c92994bb9f667dd3e74581bf2aa89f24e1d56123ab8203f4567aba0", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo podemos hablar.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Natural clear speech.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo podemos hablar.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "5397306abe3220a216c90c3f0f85663807f0ead62a1c12401ae6596eda3a4a2d", "generation_seconds_batch": 14.724913107929751, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7089178562164307, "target_z": 1.5312499675760252, "ce": 5.323658466339111, "genuineness": 4.319066047668457, "identity": 0.49520036578178406, "hypothesis": "Tu mensaje llegó después de que me fuera, lo leí dos veces y guardé el teléfono. Dime cuándo podemos hablar.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a5263b5b3c92994bb9f667dd3e74581bf2aa89f24e1d56123ab8203f4567aba0.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "a584a21d2dc643fcd39807965a147b3aa4f2630a6f0f4a199ff09adbbc6076f9", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ca009c8f5477768f6fb7001551c6ce31a5fc279b0207958ba9c53b11d7aa3d13", "generation_seconds_batch": 24.13006234099157, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4934933185577393, "target_z": 1.3749999567680335, "ce": 6.086154937744141, "genuineness": 4.344170093536377, "identity": 0.4327126741409302, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a584a21d2dc643fcd39807965a147b3aa4f2630a6f0f4a199ff09adbbc6076f9.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "a632021ba40d64bcb635e628ba4ef6b1a2c486e459bc06d2b376c5da7a941a54", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ad5600c53ce251b7f75cac8198b18a90d3ad5a103ef5572032d9618b9643a3a3", "generation_seconds_batch": 17.78372845100239, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6048672199249268, "target_z": 0.730468748649001, "ce": 6.316492557525635, "genuineness": 4.532451152801514, "identity": 0.6537661552429199, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a632021ba40d64bcb635e628ba4ef6b1a2c486e459bc06d2b376c5da7a941a54.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "a678c79ef23f3e498715593b9a88063f4ff4bb2c730f5efa9d8aea7f0e0bcc88", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "913c1cba0e4da184af4140d0e84351068fcfdb268a1b1d2daccc571537964859", "generation_seconds_batch": 21.956297968048602, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.312106609344482, "target_z": 1.9687500324239748, "ce": 5.593709468841553, "genuineness": 4.369274139404297, "identity": 0.6920340061187744, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a678c79ef23f3e498715593b9a88063f4ff4bb2c730f5efa9d8aea7f0e0bcc88.mp3" }, { "id": "heldout-es-4", "language": "es", "sentences": [ "Escribí el número en el sobre.", "Léelo bien... y comprueba las dos últimas cifras.", "Llámame cuando tengas un momento." ], "text": "Escribí el número en el sobre. Léelo bien... y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-4", "key": "a6a371f0f8cf8463e353041f23f8dbff0eef466bd407781fac32b90fe4b45447", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Escribí el número en el sobre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Llámame cuando tengas un momento.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Escribí el número en el sobre.\n(Natural clear speech.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Natural clear speech.) [3.2 seconds duration] Llámame cuando tengas un momento.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "29adbad85cde2a9a01c3afad2739b6e4dbe52c5e1cfeb6ea29a108c7fa03a2af", "generation_seconds_batch": 24.3650322050089, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.547349452972412, "target_z": 1.4140624594700315, "ce": 5.383228778839111, "genuineness": 4.331617832183838, "identity": 0.46691983938217163, "hypothesis": "Escribí el número en el sobre, léelo bien y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a6a371f0f8cf8463e353041f23f8dbff0eef466bd407781fac32b90fe4b45447.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "a6b69415e68026b9d0336a663865b919f9950cb03f263c0b79b2515ca36b6df2", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "1286b5e43b746da25242111dda5c02724ab0f4519791581c0979831193897afc", "generation_seconds_batch": 18.00180448696483, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7950878143310547, "target_z": 1.5937500756559413, "ce": 5.847874641418457, "genuineness": 4.519899368286133, "identity": 0.41505348682403564, "hypothesis": "I found the little box. It was beside the old window. We can look inside.", "wer": 0.0625, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a6b69415e68026b9d0336a663865b919f9950cb03f263c0b79b2515ca36b6df2.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "a6c4399575bd9170e71ea7bb56eecb5a1174bbdaab861266720b4a8b3387868a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "2015503044781d42939f309cd105e473b67eb08e11ee6b91f0ff6a5bea254d25", "generation_seconds_batch": 16.601274695945904, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9028000831604004, "target_z": 1.6718750810599372, "ce": 5.816103935241699, "genuineness": 4.6956281661987305, "identity": 0.5978912711143494, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a6c4399575bd9170e71ea7bb56eecb5a1174bbdaab861266720b4a8b3387868a.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "a70879c36d70a573e2c16073b7131d76c22fcb3930a75f66f6660394589f2c2b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f88d1abdf4d884be85bb7af7b9319c215e4c940fb04ae3eb6d54ac075606563f", "generation_seconds_batch": 23.775579878943972, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.159585475921631, "target_z": 1.1328125783579392, "ce": 6.300607204437256, "genuineness": 4.557555198669434, "identity": 0.788365364074707, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a70879c36d70a573e2c16073b7131d76c22fcb3930a75f66f6660394589f2c2b.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "a7d5905405bdbb45be0e74d10004c610cf90d8782f90aecd0964cf4eee19d29c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f39f42f303c84ced7431ca5e6dc9c600ebca1fac491719476e0c897d61021481", "generation_seconds_batch": 25.89438219089061, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.89202880859375, "target_z": 1.6640625459339644, "ce": 6.125868320465088, "genuineness": 4.758388519287109, "identity": 0.6470468640327454, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a7d5905405bdbb45be0e74d10004c610cf90d8782f90aecd0964cf4eee19d29c.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "a8411e65eb9fe00c884749943112d0b0563bbcc751774dd7f091ede30a17fef9", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c3183fc44496ab747a06cb86f2cb29afea4bf477b6c82a42a9f6e29f6e2164f6", "generation_seconds_batch": 23.668342157965526, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.577939033508301, "target_z": 0.7109374108340691, "ce": 6.284721851348877, "genuineness": 4.632867813110352, "identity": 0.72904372215271, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a8411e65eb9fe00c884749943112d0b0563bbcc751774dd7f091ede30a17fef9.mp3" }, { "id": "heldout-nl-4", "language": "nl", "sentences": [ "Ik schreef het nummer op de envelop.", "Lees het goed... en controleer de laatste twee cijfers.", "Bel me wanneer je even tijd hebt." ], "text": "Ik schreef het nummer op de envelop. Lees het goed... en controleer de laatste twee cijfers. Bel me wanneer je even tijd hebt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-4", "key": "a8bd001b1fcafd2f49b65d73e6ed12567e4c9664909a4d08700c9febfd977540", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "76e83843a4ce2a6c9048db6a7ddaeaf32d60f23d4aede3bb2a022bc75a560169", "generation_seconds_batch": 20.571306576952338, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5906479358673096, "target_z": -0.730468748649001, "ce": 6.252951145172119, "genuineness": 4.15588903427124, "identity": 0.7747451066970825, "hypothesis": "Ik schreef het nummer op de enveloop. Lees het goed en controleer de laatste twee cijfers. Bel me wanneer je even tijd hebt.", "wer": 0.043478260869565216, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a8bd001b1fcafd2f49b65d73e6ed12567e4c9664909a4d08700c9febfd977540.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "a8c10fec91ff5b29280701df8d707eb38f260362eaebcf34998a501b116c4fc9", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8e7a6015e8f8d69a6a008afc1206a4d7b59736be8c147c0357129db9760f57db", "generation_seconds_batch": 25.275906038004905, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4934933185577393, "target_z": 1.3749999567680335, "ce": 5.414999485015869, "genuineness": 4.268857479095459, "identity": 0.34084707498550415, "hypothesis": "La puerta sigue abierta. He dejado tu abrijo sobre la silla. Dime cuándo estás listo.", "wer": 0.06666666666666667, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a8c10fec91ff5b29280701df8d707eb38f260362eaebcf34998a501b116c4fc9.mp3" }, { "id": "heldout-en-3", "language": "en", "sentences": [ "Someone moved the little blue chair.", "It used to stand here... beside the tall bookshelf.", "We can put it back after lunch." ], "text": "Someone moved the little blue chair. It used to stand here... beside the tall bookshelf. We can put it back after lunch.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-3", "key": "a8ca64bde8999fbe8ff1bb79d59535add6e785f1455e701e0ffb52f0623087dd", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Someone moved the little blue chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can put it back after lunch.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Someone moved the little blue chair.\n(Natural clear speech.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Natural clear speech.) [3.2 seconds duration] We can put it back after lunch.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7a0db888b3c86e863f2431938512fe0fd2ca64d2d8f4d014348a78bd6b389829", "generation_seconds_batch": 27.698526351014152, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2888400554656982, "target_z": 1.2265624810860147, "ce": 5.990842819213867, "genuineness": 4.325342178344727, "identity": 0.3791166841983795, "hypothesis": "Someone moved the little blue chair. It used to stand here beside the tall bookshelf. We can put it back after lunch.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a8ca64bde8999fbe8ff1bb79d59535add6e785f1455e701e0ffb52f0623087dd.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "a8cfe83241f8290cc42a02283c7395159fdc88bcb9ffb2256bb9036c77c202b4", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "666bcf9c8388d3b47f80b2c45b15b381636fe8907cb5787da800f15bd30e48ab", "generation_seconds_batch": 17.6027803060133, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5042645931243896, "target_z": 1.3828124918940063, "ce": 5.605623245239258, "genuineness": 4.193545341491699, "identity": 0.4603448808193207, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinräumen.", "wer": 0.0625, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a8cfe83241f8290cc42a02283c7395159fdc88bcb9ffb2256bb9036c77c202b4.mp3" }, { "id": "heldout-nl-3", "language": "nl", "sentences": [ "Iemand heeft het kleine blauwe stoeltje verplaatst.", "Het stond hier... naast de hoge boekenkast.", "We zetten het na de lunch terug." ], "text": "Iemand heeft het kleine blauwe stoeltje verplaatst. Het stond hier... naast de hoge boekenkast. We zetten het na de lunch terug.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-3", "key": "a8e16ddf26a307ebf1855b3b3b0b5fd6faae1d543f5d02689986b8e0675f2311", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We zetten het na de lunch terug.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Natural clear speech.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Natural clear speech.) [3.2 seconds duration] We zetten het na de lunch terug.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "091d0059c7d035dcc3d3f01099e8a19572c379f6138bcbfcfc584b588fccd085", "generation_seconds_batch": 17.493517817929387, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.879533529281616, "target_z": 0.929687529721977, "ce": 6.292664527893066, "genuineness": 4.6705241203308105, "identity": 0.7805159091949463, "hypothesis": "Iemand heeft het kleine blauwe stoeltje verplaatst. Het stond hier naast de hoge boekenkast. We zetten het na de lunch terug.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a8e16ddf26a307ebf1855b3b3b0b5fd6faae1d543f5d02689986b8e0675f2311.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "a9147e9705255ec4386e233413118c70706ddf345c4c06a34fc569ff63d94e57", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "69406ed90a3d0ece968709f01b61a46273b333c43263355a8fc5fbae89e8d6ab", "generation_seconds_batch": 24.140849750023335, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.437913179397583, "target_z": 0.609374972980021, "ce": 6.181467056274414, "genuineness": 4.545003414154053, "identity": 0.6422017216682434, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a9147e9705255ec4386e233413118c70706ddf345c4c06a34fc569ff63d94e57.mp3" }, { "id": "heldout-fr-0", "language": "fr", "sentences": [ "Ton message est arrivé après mon départ.", "Je l’ai relu... puis j’ai rangé mon téléphone.", "Dis-moi quand nous pourrons parler." ], "text": "Ton message est arrivé après mon départ. Je l’ai relu... puis j’ai rangé mon téléphone. Dis-moi quand nous pourrons parler.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-0", "key": "a94b7d5da78ff217f8e5a4f7ae4520900bef0ad764d0e4ee898e69de10c28aea", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e06551016a0ea564d1eb1b7cb825e5264e74d46f614e6473f34bec6c4e2e8493", "generation_seconds_batch": 15.886211984907277, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4321500062942505, "target_z": -0.12011716554626703, "ce": 6.252951145172119, "genuineness": 4.381826400756836, "identity": 0.7312057018280029, "hypothesis": "Ton message est arrivé après mon départ. Je l'ai relu, puis j'ai rangé mon téléphone. Dis-moi, quand nous pourrons parler ?", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a94b7d5da78ff217f8e5a4f7ae4520900bef0ad764d0e4ee898e69de10c28aea.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "a9a79c294637915105081e9e2fd3fe84fdc8a05236f3fab1cb47fdfba7f3a101", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2e580c84c91ed8d5bac06a06f0f24d3c267d56ccb5430ee94f4b823a37557e51", "generation_seconds_batch": 24.47754904802423, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1703567504882812, "target_z": 1.140625113483912, "ce": 6.1020402908325195, "genuineness": 4.545003414154053, "identity": 0.8124060034751892, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/a9a79c294637915105081e9e2fd3fe84fdc8a05236f3fab1cb47fdfba7f3a101.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "ab2e75517dd382ce8f79cc4b6636a777fae57d1248f710d399812a789d228d01", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e94aeaca92ace3880109062585ff972b39a378f09057d879c813d85e3bdb2f0e", "generation_seconds_batch": 24.829344674013555, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9226183891296387, "target_z": 0.960937497298002, "ce": 5.700935363769531, "genuineness": 4.162164688110352, "identity": 0.41176116466522217, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás lista.", "wer": 0.06666666666666667, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ab2e75517dd382ce8f79cc4b6636a777fae57d1248f710d399812a789d228d01.mp3" }, { "id": "heldout-en-0", "language": "en", "sentences": [ "Your message arrived after I had left.", "I read it twice... before putting the phone away.", "Let me know when we can talk." ], "text": "Your message arrived after I had left. I read it twice... before putting the phone away. Let me know when we can talk.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-0", "key": "ab7a94299f968552a7b966fa3391724b652d8e20483888a7767196f29a7f4a69", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Your message arrived after I had left.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Let me know when we can talk.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Your message arrived after I had left.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Let me know when we can talk.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1f7ff862b520b4eb96f5b6b8faa2432c8ae847a8c5e7b96728f06176cf0cd8ef", "generation_seconds_batch": 16.32647258997895, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7398030757904053, "target_z": 0.10302730980615132, "ce": 6.41180419921875, "genuineness": 4.13078498840332, "identity": 0.3442877531051636, "hypothesis": "Your message arrived after I had left. I read it twice before putting the phone away. Let me know when we can talk.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ab7a94299f968552a7b966fa3391724b652d8e20483888a7767196f29a7f4a69.mp3" }, { "id": "heldout-en-0", "language": "en", "sentences": [ "Your message arrived after I had left.", "I read it twice... before putting the phone away.", "Let me know when we can talk." ], "text": "Your message arrived after I had left. I read it twice... before putting the phone away. Let me know when we can talk.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-0", "key": "abdda4eff0137d3fbc50ced0b03b2bf24b474fa08f202727abce8a4b382be990", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Your message arrived after I had left.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Let me know when we can talk.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Your message arrived after I had left.\n(Natural clear speech.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Natural clear speech.) [3.2 seconds duration] Let me know when we can talk.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ae6de572246e7451525049251bd859da06aa143a019dcd041369762bd8dad589", "generation_seconds_batch": 23.63622226798907, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9710888862609863, "target_z": 0.9960937324370136, "ce": 5.522225379943848, "genuineness": 4.356721878051758, "identity": 0.5164270401000977, "hypothesis": "Your message arrived after I had left. I read it twice before putting the phone away. Let me know when we can talk.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/abdda4eff0137d3fbc50ced0b03b2bf24b474fa08f202727abce8a4b382be990.mp3" }, { "id": "heldout-de-2", "language": "de", "sentences": [ "Die Straße führt hinter der Brücke nach links.", "Geh weiter... bis zu dem kleinen Holztor.", "Ich warte auf der anderen Seite." ], "text": "Die Straße führt hinter der Brücke nach links. Geh weiter... bis zu dem kleinen Holztor. Ich warte auf der anderen Seite.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-2", "key": "abf3701a5561b37fd3ef0c077c053938f05d317b72c948d38805d0a22d6ede9c", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Natural clear speech.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Natural clear speech.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "17e96bdb5383ada0fc8169ad373f15ea71eb4d75226d4e2772c1a434c55d8635", "generation_seconds_batch": 15.984086344949901, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.010512351989746, "target_z": 1.750000086463933, "ce": 5.855817794799805, "genuineness": 4.545003414154053, "identity": 0.5539225339889526, "hypothesis": "Die Straße führt hinter der Brücke nach links. Geh weiter bis zu dem kleinen Holztor. Ich warte auf der anderen Seite.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/abf3701a5561b37fd3ef0c077c053938f05d317b72c948d38805d0a22d6ede9c.mp3" }, { "id": "heldout-fr-1", "language": "fr", "sentences": [ "Les clés sont sur la table de la cuisine.", "J’ai regardé dans le tiroir... puis près de la fenêtre.", "Prends-les quand tu partiras." ], "text": "Les clés sont sur la table de la cuisine. J’ai regardé dans le tiroir... puis près de la fenêtre. Prends-les quand tu partiras.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-1", "key": "acca1b1145c297e14bce8636e547201a201ce4ab4d34d6410099cb07209a9e00", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Prends-les quand tu partiras.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Natural clear speech.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Prends-les quand tu partiras.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a71877facf04a79014a7d11d0ca05e809ae62fb5b3a86e5ec92f32b3b6f146b8", "generation_seconds_batch": 18.414642314077355, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.752002716064453, "target_z": 1.5624999351520503, "ce": 5.720792293548584, "genuineness": 4.607763767242432, "identity": 0.728340744972229, "hypothesis": "Les clés sont sur la table de la cuisine j'ai regardé dans le tiroir, puis près de la fenêtre. Prends-les quand tu partiras.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/acca1b1145c297e14bce8636e547201a201ce4ab4d34d6410099cb07209a9e00.mp3" }, { "id": "heldout-de-4", "language": "de", "sentences": [ "Ich schrieb die Nummer auf den Umschlag.", "Lies sie genau... und prüfe die letzten beiden Ziffern.", "Ruf an, wenn du Zeit hast." ], "text": "Ich schrieb die Nummer auf den Umschlag. Lies sie genau... und prüfe die letzten beiden Ziffern. Ruf an, wenn du Zeit hast.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-4", "key": "ad228a28e6bacfa5231cccb28b57d7bc38c96656409bcf4739c26aeafd127226", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Natural clear speech.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Natural clear speech.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "de0fb2fb0920faeacd06735a55787371386d3f0a1d1e678d101ca31ced05b3b3", "generation_seconds_batch": 16.94337453308981, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4396371841430664, "target_z": 1.3359374540660356, "ce": 5.847874641418457, "genuineness": 4.482243061065674, "identity": 0.5835775136947632, "hypothesis": "Ich schrieb die Nummer auf den Umschlag, ließ sie genau und prüfe die letzten beiden Ziffern. Ruf an, wenn du Zeit hast.", "wer": 0.045454545454545456, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ad228a28e6bacfa5231cccb28b57d7bc38c96656409bcf4739c26aeafd127226.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "ad381820d1e8f0f5f13d7666060f89299626f78bc130ce389378dc214f43d098", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d302044578d2c40b22ed3de6bd0ca67a18b519129fb9b8f796bc7f23730cd9f1", "generation_seconds_batch": 22.975548913003877, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.051873207092285, "target_z": 1.0546875729539433, "ce": 6.181467056274414, "genuineness": 4.532451152801514, "identity": 0.5284464955329895, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ad381820d1e8f0f5f13d7666060f89299626f78bc130ce389378dc214f43d098.mp3" }, { "id": "heldout-es-0", "language": "es", "sentences": [ "Tu mensaje llegó después de que me fuera.", "Lo leí dos veces... y guardé el teléfono.", "Dime cuándo podemos hablar." ], "text": "Tu mensaje llegó después de que me fuera. Lo leí dos veces... y guardé el teléfono. Dime cuándo podemos hablar.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-0", "key": "ad49ba9e5ebe1244c5f29f89ad6092cac75138b1378f4607f378c724074012f3", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo podemos hablar.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Natural clear speech.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo podemos hablar.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "71f1663edfd1cdcb4f2391f8d73bb84ca3d4e37bd8ba7dbbaee22b55af7f1a1f", "generation_seconds_batch": 23.700036367052235, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4719510078430176, "target_z": 1.359375059443954, "ce": 5.414999485015869, "genuineness": 4.306513786315918, "identity": 0.4151695966720581, "hypothesis": "Tu mensaje llegó después de que me fuera. Lo leí dos veces y guardé el teléfono. Dime cuándo podemos hablar.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ad49ba9e5ebe1244c5f29f89ad6092cac75138b1378f4607f378c724074012f3.mp3" }, { "id": "heldout-nl-1", "language": "nl", "sentences": [ "De sleutels liggen op de keukentafel.", "Ik keek in de la... en daarna naast het raam.", "Neem ze mee als je weggaat." ], "text": "De sleutels liggen op de keukentafel. Ik keek in de la... en daarna naast het raam. Neem ze mee als je weggaat.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-1", "key": "ad593c84e0cce3b8f0668a24c611337fb9a537d23d1f025fbebe62edcb7be6bf", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Neem ze mee als je weggaat.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Natural clear speech.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Natural clear speech.) [3.2 seconds duration] Neem ze mee als je weggaat.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "17d727b6c6d2983c45c43bf600a0e42476a549a3063521ba0798a83eadbb6860", "generation_seconds_batch": 15.059516415931284, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.838172674179077, "target_z": 1.6250000432319665, "ce": 5.681078910827637, "genuineness": 4.419482707977295, "identity": 0.326033353805542, "hypothesis": "De sleutels liggen op de keukentafel. Ik keek in de laad. En daarna, naast het raam, neem ze mee als je weggaat.", "wer": 0.045454545454545456, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ad593c84e0cce3b8f0668a24c611337fb9a537d23d1f025fbebe62edcb7be6bf.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "ad66948af826e9f6a2f6e7b54086bdcc472cf2183398024fec4d7d57e5d234ac", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "cf6169777eb8c4c9762e9ee02db47e3708fe0ed979786971185c749065761dae", "generation_seconds_batch": 25.912818546989, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3732857704162598, "target_z": 0.5624999351520503, "ce": 6.086154937744141, "genuineness": 4.607763767242432, "identity": 0.4270932674407959, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ad66948af826e9f6a2f6e7b54086bdcc472cf2183398024fec4d7d57e5d234ac.mp3" }, { "id": "heldout-nl-0", "language": "nl", "sentences": [ "Je bericht kwam nadat ik vertrokken was.", "Ik las het twee keer... en legde mijn telefoon weg.", "Zeg maar wanneer we kunnen praten." ], "text": "Je bericht kwam nadat ik vertrokken was. Ik las het twee keer... en legde mijn telefoon weg. Zeg maar wanneer we kunnen praten.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-0", "key": "adcc0121d6c11de6953f8243f84ad890832f75283c96342c1db246fddc443438", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5ad33010a6cf0748c871496bc72721d0c7f568215e496b3bc643b5b27ba9f417", "generation_seconds_batch": 17.577438231906854, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7953012585639954, "target_z": -0.5820312297350158, "ce": 6.213237762451172, "genuineness": 4.482243061065674, "identity": 0.6237586736679077, "hypothesis": "Je bericht kwam nadat ik vertrokken was. Ik las het twee keer en legde mijn telefoon weg. Zeg maar wanneer we kunnen praten.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/adcc0121d6c11de6953f8243f84ad890832f75283c96342c1db246fddc443438.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "adcd817f9fc08657b06533df2025892e7b7f766035d30f8b04e3dc10f0bcdb32", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "069c9c992ad3cf03fea166275d8d646a8fc8c4603e11a46d82dabea95d276e4c", "generation_seconds_batch": 23.06798124301713, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4719510078430176, "target_z": 1.359375059443954, "ce": 5.824047088623047, "genuineness": 4.319066047668457, "identity": 0.6787219047546387, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/adcd817f9fc08657b06533df2025892e7b7f766035d30f8b04e3dc10f0bcdb32.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "adf8373f09f151ad0d62b5dc80bb0b43aa6830d995eb3d419eed2339a7ba14f6", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "73e8040411f986c1f2c74b54f0a5aace3c306b25a04b1e19b8ce82a9b55460ea", "generation_seconds_batch": 22.20864488906227, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9764745235443115, "target_z": 1.0, "ce": 6.229123115539551, "genuineness": 4.268857479095459, "identity": 0.3501511812210083, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/adf8373f09f151ad0d62b5dc80bb0b43aa6830d995eb3d419eed2339a7ba14f6.mp3" }, { "id": "heldout-en-2", "language": "en", "sentences": [ "The road turns left behind the bridge.", "Keep going... until you reach the small wooden gate.", "I will meet you on the other side." ], "text": "The road turns left behind the bridge. Keep going... until you reach the small wooden gate. I will meet you on the other side.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-2", "key": "ae3130a5db36c2790a31e23135fde257111c56fa9c5b8061c1f3e51a8cb61258", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The road turns left behind the bridge.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I will meet you on the other side.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The road turns left behind the bridge.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I will meet you on the other side.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "435e2e932c2c52fe016abca1c71f7f0c0e26e555ec1482c4db6bfc4f83e59a5f", "generation_seconds_batch": 16.38473025301937, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.592371940612793, "target_z": -0.003906267562986381, "ce": 6.348263263702393, "genuineness": 4.005263805389404, "identity": 0.30984944105148315, "hypothesis": "The road turns left behind the bridge. Keep going until you reach the small wooden gate. I will meet you on the other side.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ae3130a5db36c2790a31e23135fde257111c56fa9c5b8061c1f3e51a8cb61258.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "ae4e4de6200fef1d21a8c5e98354466b3652831c177e1b12099ba0a95c989364", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "0b23b04d35228a7e0189bf5fb7ac38ac19cd2afe2d9d5c24b58d9f9ea6cc15a0", "generation_seconds_batch": 23.315962848020718, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5258069038391113, "target_z": 1.398437389218086, "ce": 5.804190158843994, "genuineness": 4.319066047668457, "identity": 0.31353241205215454, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ae4e4de6200fef1d21a8c5e98354466b3652831c177e1b12099ba0a95c989364.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "ae4e95c88fc8b67f9d545d6586e3ea75492b76c9e3f62cf87ca409c6fe2dacd2", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ae02773bd576eb7523492a25ca4185290cd4c8b13f057accb024ad07abcfdf70", "generation_seconds_batch": 17.78372845100239, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5910255908966064, "target_z": -0.004882791221766499, "ce": 6.268836498260498, "genuineness": 4.457139015197754, "identity": 0.4822843372821808, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ae4e95c88fc8b67f9d545d6586e3ea75492b76c9e3f62cf87ca409c6fe2dacd2.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "aece3b3b2d7da239a419b3c03db36d444f795433201f714aff766c1a4fbab0f7", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3e3d1baea724326b3cdc4ad872a53351d55ab7c33cd3bfcb03f8b2553c0303c7", "generation_seconds_batch": 23.278622932964936, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5581207275390625, "target_z": 1.421874994596004, "ce": 6.149696350097656, "genuineness": 4.432034492492676, "identity": 0.5850257277488708, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/aece3b3b2d7da239a419b3c03db36d444f795433201f714aff766c1a4fbab0f7.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "af1647128ba56b0a236142a62292cb7ebb58bb12ea14e60c908816f4d473b472", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0207c191764e3378a8fcd333fe089be4cc99017c3e82a6e0e3cc414d57394860", "generation_seconds_batch": 17.252815083018504, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.95493221282959, "target_z": 0.9843751026759204, "ce": 6.229123115539551, "genuineness": 4.482243061065674, "identity": 0.6223818063735962, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/af1647128ba56b0a236142a62292cb7ebb58bb12ea14e60c908816f4d473b472.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "b0161494287ff730083ce84d02f1814a3868fa2e38d96a80f51c76c3938a0617", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1b9ae5097f50391e7701ee35f2e48f990aa1bfd701479dc5698814e54e7d9bb6", "generation_seconds_batch": 22.431560627068393, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6119768619537354, "target_z": 1.460937497298002, "ce": 5.760505676269531, "genuineness": 4.494794845581055, "identity": 0.39876121282577515, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b0161494287ff730083ce84d02f1814a3868fa2e38d96a80f51c76c3938a0617.mp3" }, { "id": "heldout-es-1", "language": "es", "sentences": [ "Las llaves están en la mesa de la cocina.", "Miré en el cajón... y luego junto a la ventana.", "Llévatelas cuando te vayas." ], "text": "Las llaves están en la mesa de la cocina. Miré en el cajón... y luego junto a la ventana. Llévatelas cuando te vayas.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-1", "key": "b023dc9072901d215577cbdc77a3fc74234f800a09404eabde41fd265fccd73d", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Llévatelas cuando te vayas.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Natural clear speech.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Natural clear speech.) [3.2 seconds duration] Llévatelas cuando te vayas.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e804f0b6818c84d6a05587cbac06e2d338d20952947eec35d62abbfffb00d9b2", "generation_seconds_batch": 14.916390712023713, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.601205587387085, "target_z": 1.4531249621720292, "ce": 5.7406487464904785, "genuineness": 4.400654315948486, "identity": 0.5064095258712769, "hypothesis": "Las llaves están en la mesa de la cocina. Miren el cajón y luego junto a la ventana, llévatelas cuando te vayas.", "wer": 0.08695652173913043, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b023dc9072901d215577cbdc77a3fc74234f800a09404eabde41fd265fccd73d.mp3" }, { "id": "heldout-en-4", "language": "en", "sentences": [ "I wrote the number on the envelope.", "Read it carefully... and check the last two digits.", "Call me when you have a moment." ], "text": "I wrote the number on the envelope. Read it carefully... and check the last two digits. Call me when you have a moment.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-4", "key": "b0b4dc01440c340938a223fb70cc98bf781fc102db9f99d1f1f8a7c329ce4e0e", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I wrote the number on the envelope.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Call me when you have a moment.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I wrote the number on the envelope.\n(Natural clear speech.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Natural clear speech.) [3.2 seconds duration] Call me when you have a moment.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "b5e4298e12337a6cfc6d9a294ec6d28d6f38ed6ef8d2a95d63b8b5f248be2ab4", "generation_seconds_batch": 16.308117015985772, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.874147891998291, "target_z": 0.9257812621589906, "ce": 5.927301406860352, "genuineness": 3.8922953605651855, "identity": 0.5456598997116089, "hypothesis": "I wrote the number on the envelope. Read it carefully and check the last two digits. Call me when you have", "wer": 0.08695652173913043, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b0b4dc01440c340938a223fb70cc98bf781fc102db9f99d1f1f8a7c329ce4e0e.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "b0bd10ec5a7857855abb9c4eac3c98e1bdb0208540167bdb41f142ec683f3fc0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0d1ae5492eca3376027d3c067ca6e4dd2123316d24c163263004bf98edfcbda0", "generation_seconds_batch": 22.30017531698104, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1272716522216797, "target_z": 1.1093749729800209, "ce": 6.1894097328186035, "genuineness": 4.268857479095459, "identity": 0.3434520959854126, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b0bd10ec5a7857855abb9c4eac3c98e1bdb0208540167bdb41f142ec683f3fc0.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "b118c6e020cb61043f275793af6fd3eb82bd219e47dbb9624a860379ba0c0cec", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7b83f167e4e2366cabf71e5fdf11ac9f2f488c5171201052c8544af039cf6476", "generation_seconds_batch": 24.093101347913034, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2888400554656982, "target_z": 1.2265624810860147, "ce": 5.935244083404541, "genuineness": 4.444586753845215, "identity": 0.5862826108932495, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar, wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b118c6e020cb61043f275793af6fd3eb82bd219e47dbb9624a860379ba0c0cec.mp3" }, { "id": "heldout-de-4", "language": "de", "sentences": [ "Ich schrieb die Nummer auf den Umschlag.", "Lies sie genau... und prüfe die letzten beiden Ziffern.", "Ruf an, wenn du Zeit hast." ], "text": "Ich schrieb die Nummer auf den Umschlag. Lies sie genau... und prüfe die letzten beiden Ziffern. Ruf an, wenn du Zeit hast.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-4", "key": "b1206621569f452e8361a53c64d888cdc7ff3d65ece6e3cfc3576a4404e0a572", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Natural clear speech.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Natural clear speech.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1d206dc91fddb084a56a5c55301bbb1a36ee8e3cf4ecaeabd80cc4d0d670efe1", "generation_seconds_batch": 16.163580458029173, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.601205587387085, "target_z": 1.4531249621720292, "ce": 5.903473377227783, "genuineness": 4.469690799713135, "identity": 0.5917824506759644, "hypothesis": "Ich schrieb die Nummer auf den Umschlag, lies sie genau und prüfe die letzten beiden Ziffern. Ruf an, wenn du Zeit hast.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b1206621569f452e8361a53c64d888cdc7ff3d65ece6e3cfc3576a4404e0a572.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "b1208d63a7463e4bb0495dd1384077ef02499bc0b587cd6c634b799d9f7ced9a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a01a6c91db9e2d652ed313b0d2fef6b02b655a3bcfcec4a1a95089627f11503d", "generation_seconds_batch": 22.47322924400214, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2996113300323486, "target_z": 1.2343750162119875, "ce": 6.252951145172119, "genuineness": 4.595211505889893, "identity": 0.5866469740867615, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b1208d63a7463e4bb0495dd1384077ef02499bc0b587cd6c634b799d9f7ced9a.mp3" }, { "id": "heldout-fr-1", "language": "fr", "sentences": [ "Les clés sont sur la table de la cuisine.", "J’ai regardé dans le tiroir... puis près de la fenêtre.", "Prends-les quand tu partiras." ], "text": "Les clés sont sur la table de la cuisine. J’ai regardé dans le tiroir... puis près de la fenêtre. Prends-les quand tu partiras.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-1", "key": "b13b3f31ed8a598ed0c7ff09ef0026bb03b1d767b8b7cde6700b5d61816bbf51", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Prends-les quand tu partiras.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Natural clear speech.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Prends-les quand tu partiras.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d7fce7b51af11ac5f6934cfe29f11c7012a4651ffd385507197f0112b8bdff30", "generation_seconds_batch": 15.476682175998576, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5365781784057617, "target_z": 1.4062499243440587, "ce": 5.5659098625183105, "genuineness": 4.419482707977295, "identity": 0.7033525705337524, "hypothesis": "Les clés sont sur la table de la cuisine. J'ai regardé dans le tiroir, puis près de la fenêtre. Prends-les quand tu partiras.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b13b3f31ed8a598ed0c7ff09ef0026bb03b1d767b8b7cde6700b5d61816bbf51.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "b14a3e5518ef2283387b00eba49cdb6176e5c3e6654f04f17c753442ddc92e01", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "db8b4159ec0686ddd207212871c54dadb4d0bce142d3c6cb3fd4d50618592873", "generation_seconds_batch": 22.83836459298618, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.665832996368408, "target_z": 1.5, "ce": 5.895530700683594, "genuineness": 4.256305694580078, "identity": 0.46602192521095276, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b14a3e5518ef2283387b00eba49cdb6176e5c3e6654f04f17c753442ddc92e01.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "b19bc7b83b4f02c462be66c42ef15c935ca73e9c2a17b88fd9d65de3a1dfdb08", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "0a1b2be0a782cb1358a3222b7ecc295f152b7c341aed8eb7d059dac4dc3c81df", "generation_seconds_batch": 17.800323338015005, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.279792785644531, "target_z": 1.9453124270460567, "ce": 5.927301406860352, "genuineness": 4.7332844734191895, "identity": 0.7096915245056152, "hypothesis": "J'ai trouvé la petite boîte elle était près de la vieille fenêtre nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b19bc7b83b4f02c462be66c42ef15c935ca73e9c2a17b88fd9d65de3a1dfdb08.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "b2003a4f77746412d8f53323de161312e43942004273b61980e1650ce5a0697d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "9d2666717705f22af349a7b8aec47cfbcd9426fc249846eb1212400374b7c6bc", "generation_seconds_batch": 23.68443738296628, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0303306579589844, "target_z": 1.039062502701998, "ce": 6.205295085906982, "genuineness": 4.8588056564331055, "identity": 0.8231817483901978, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b2003a4f77746412d8f53323de161312e43942004273b61980e1650ce5a0697d.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "b2539ceddc65a7f376c2861575f5cee8a64dba1af8f721e300604081ed1055e4", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "47c625f01ff34776e949870f68025b5ece66c49fe7b686fe692c51db5a83108c", "generation_seconds_batch": 21.150289794895798, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0303306579589844, "target_z": 1.039062502701998, "ce": 6.443574905395508, "genuineness": 4.482243061065674, "identity": 0.39018717408180237, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b2539ceddc65a7f376c2861575f5cee8a64dba1af8f721e300604081ed1055e4.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "b2c1a285af63a4713209eac5b2477cc9ccf812020b970447ce88690ae4ebd405", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "03ec40b2bbdb9001dd3010604efe7976f5117966c099949f9dba7f777fcf63c4", "generation_seconds_batch": 24.463273369008675, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0626444816589355, "target_z": 1.0625001080799161, "ce": 6.0146708488464355, "genuineness": 4.6203155517578125, "identity": 0.7887152433395386, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b2c1a285af63a4713209eac5b2477cc9ccf812020b970447ce88690ae4ebd405.mp3" }, { "id": "heldout-nl-0", "language": "nl", "sentences": [ "Je bericht kwam nadat ik vertrokken was.", "Ik las het twee keer... en legde mijn telefoon weg.", "Zeg maar wanneer we kunnen praten." ], "text": "Je bericht kwam nadat ik vertrokken was. Ik las het twee keer... en legde mijn telefoon weg. Zeg maar wanneer we kunnen praten.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-0", "key": "b2f585f98fe4dc626b50f855d93d400d86301513df55b13657f411dca4a1a717", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Natural clear speech.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e8c75708496acfce2430b8d1bb51bb754c9fcab076c60a3f17c0d5d8e097aa02", "generation_seconds_batch": 16.65428311296273, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.773545265197754, "target_z": 1.578125005403996, "ce": 6.046441555023193, "genuineness": 4.607763767242432, "identity": 0.1693456470966339, "hypothesis": "Je bericht kwam nadat ik vertrokken was. Ik las het twee keer en legde mijn telefoon weg. Zeg maar wanneer we kunnen praten.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b2f585f98fe4dc626b50f855d93d400d86301513df55b13657f411dca4a1a717.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "b31d38fa7f484ada69c9d576635cc31145f78e2f43f248c0d5b3f0e407041a64", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "151d1c24a394149a373f82b4d846121c81910b36dfebe34d28973ca4e864dd0d", "generation_seconds_batch": 15.820469844038598, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1703567504882812, "target_z": 1.140625113483912, "ce": 6.125868320465088, "genuineness": 4.344170093536377, "identity": 0.7040539979934692, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b31d38fa7f484ada69c9d576635cc31145f78e2f43f248c0d5b3f0e407041a64.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "b36eb2373a4f9ec240545973aaf9038d8cdef09dbbeb73224033a3571c9b7787", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f69d2e2dd5abc811c16303c2ad51927c77c81e349f9ce6b10c14f4c97c9c8025", "generation_seconds_batch": 15.820469844038598, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3319249153137207, "target_z": 1.25781244866204, "ce": 6.1894097328186035, "genuineness": 4.570107460021973, "identity": 0.6604847311973572, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b36eb2373a4f9ec240545973aaf9038d8cdef09dbbeb73224033a3571c9b7787.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "b4000208fc266b58bb14b754820f9a3c473b7ed474e2bb1685963c17f5b8cedb", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.0, "cfg": 2.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "9ea4c1b3ab21ed37ccb1f554e03c9adbae626846183844e7e70ac21fca9be4a7", "generation_seconds_batch": 18.789165863068774, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6227481365203857, "target_z": 1.4687500324239748, "ce": 6.062326908111572, "genuineness": 4.557555198669434, "identity": 0.7781021595001221, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b4000208fc266b58bb14b754820f9a3c473b7ed474e2bb1685963c17f5b8cedb.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "b401eef0cd65409e6d5d3522802b81b9c2037083fe127807b282aeb4401ca066", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "04f477f52a7a547854635121370e0aa48e1f28415a5ee583319d98b490e8c150", "generation_seconds_batch": 22.407040906953625, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.965703248977661, "target_z": 0.9921874648740272, "ce": 6.070269584655762, "genuineness": 4.356721878051758, "identity": 0.47027653455734253, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b401eef0cd65409e6d5d3522802b81b9c2037083fe127807b282aeb4401ca066.mp3" }, { "id": "heldout-de-1", "language": "de", "sentences": [ "Die Schlüssel liegen auf dem Küchentisch.", "Ich suchte in der Schublade... dann neben dem Fenster.", "Nimm sie mit, wenn du gehst." ], "text": "Die Schlüssel liegen auf dem Küchentisch. Ich suchte in der Schublade... dann neben dem Fenster. Nimm sie mit, wenn du gehst.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-1", "key": "b429ad869b4f073af736e4e6510807542fc4f5324b0eb7f096c719d9002f4f71", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e134e43029af33d5d8232520890821bb1810c0227c373e8cfab9d1e874df3632", "generation_seconds_batch": 18.238929986022413, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.319429874420166, "target_z": 0.5234376053779183, "ce": 6.395919322967529, "genuineness": 4.406930446624756, "identity": 0.6074783802032471, "hypothesis": "Die Schlüssel liegen auf dem Küchentisch. Ich suchte in der Schublade, dann neben dem Fenster. Nimm sie mit, wenn du gehst.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b429ad869b4f073af736e4e6510807542fc4f5324b0eb7f096c719d9002f4f71.mp3" }, { "id": "heldout-nl-0", "language": "nl", "sentences": [ "Je bericht kwam nadat ik vertrokken was.", "Ik las het twee keer... en legde mijn telefoon weg.", "Zeg maar wanneer we kunnen praten." ], "text": "Je bericht kwam nadat ik vertrokken was. Ik las het twee keer... en legde mijn telefoon weg. Zeg maar wanneer we kunnen praten.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-0", "key": "b4aaa3523715483280155af578b61c38a052262906f5c5ac9c92ae0d63ce0391", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Natural clear speech.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e721e120d2addb0bcda28a712dedf75d836afa90ae3f90bd4966aa2c08da5ef5", "generation_seconds_batch": 25.183655660948716, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5365781784057617, "target_z": 1.4062499243440587, "ce": 5.780362129211426, "genuineness": 4.325342178344727, "identity": 0.2795993685722351, "hypothesis": "Je bericht kwam nadat ik vertrokken was. Ik las het twee keer en legde mijn telefoon weg. Zeg maar wanneer we kunnen praten.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b4aaa3523715483280155af578b61c38a052262906f5c5ac9c92ae0d63ce0391.mp3" }, { "id": "heldout-en-1", "language": "en", "sentences": [ "The keys are on the kitchen table.", "I checked the drawer... then looked beside the window.", "Take them with you when you leave." ], "text": "The keys are on the kitchen table. I checked the drawer... then looked beside the window. Take them with you when you leave.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-1", "key": "b4b0265d44d868b9aa7fe7c1c4ed6588c48072e5c835a1877a892854142db685", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The keys are on the kitchen table.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Take them with you when you leave.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The keys are on the kitchen table.\n(Natural clear speech.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Natural clear speech.) [3.2 seconds duration] Take them with you when you leave.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "0cf67b397f45bc963cfe6cfd58a82021823e40b2d02d7dd66082095386c05f95", "generation_seconds_batch": 17.757892654975876, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2134413719177246, "target_z": 1.1718749081320712, "ce": 6.006728172302246, "genuineness": 4.268857479095459, "identity": 0.45367830991744995, "hypothesis": "The keys are on the kitchen table. I checked the drawer, then looked beside the window. Take them with you when you leave.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b4b0265d44d868b9aa7fe7c1c4ed6588c48072e5c835a1877a892854142db685.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "b51dcc59196f2efbe6c52dee206444c13fc24ce881c525df87a0cdf678a311a9", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8041d98d08d6b26259151da37e0e22a9efe1d6dac7a452e8b87d1122c22f4b30", "generation_seconds_batch": 23.315962848020718, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5365781784057617, "target_z": 1.4062499243440587, "ce": 5.7247633934021, "genuineness": 4.444586753845215, "identity": 0.4736326336860657, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b51dcc59196f2efbe6c52dee206444c13fc24ce881c525df87a0cdf678a311a9.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "b62f7b03542466a638ad856c3f262de7f7ece5fb0a137af5b1b14b2f9a84ac7b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d00b0c4c87657ff16f09e33a80acf231c9375417e26db56691972e5d35c8443b", "generation_seconds_batch": 24.082210180000402, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0663058757781982, "target_z": 0.339843721629022, "ce": 6.2767791748046875, "genuineness": 4.570107460021973, "identity": 0.5314275026321411, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b62f7b03542466a638ad856c3f262de7f7ece5fb0a137af5b1b14b2f9a84ac7b.mp3" }, { "id": "heldout-en-3", "language": "en", "sentences": [ "Someone moved the little blue chair.", "It used to stand here... beside the tall bookshelf.", "We can put it back after lunch." ], "text": "Someone moved the little blue chair. It used to stand here... beside the tall bookshelf. We can put it back after lunch.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-3", "key": "b6422e7e5cfc7bceea08cb4e9e5c13b001e081b885ee044cc850df2cc3cda3cc", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Someone moved the little blue chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can put it back after lunch.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Someone moved the little blue chair.\n(Natural clear speech.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Natural clear speech.) [3.2 seconds duration] We can put it back after lunch.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "8e751864cc0dd2da9bcdfccb8c3d23a616c28f2d1a1bf3b4eac87d266e3dc2ca", "generation_seconds_batch": 16.48735069297254, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1488142013549805, "target_z": 1.1250000432319665, "ce": 5.959072113037109, "genuineness": 4.118232727050781, "identity": 0.42025795578956604, "hypothesis": "Someone moved the little blue chair. It used to stand here beside the tall bookshelf. We can put it back after a little bit.", "wer": 0.13636363636363635, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b6422e7e5cfc7bceea08cb4e9e5c13b001e081b885ee044cc850df2cc3cda3cc.mp3" }, { "id": "heldout-es-2", "language": "es", "sentences": [ "La carretera gira a la izquierda tras el puente.", "Sigue... hasta la pequeña puerta de madera.", "Te esperaré al otro lado." ], "text": "La carretera gira a la izquierda tras el puente. Sigue... hasta la pequeña puerta de madera. Te esperaré al otro lado.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-2", "key": "b6c5e532cb1fc50e3595a0ee36dcf6320970276d74dedd28a266768325221cc1", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Te esperaré al otro lado.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Natural clear speech.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Natural clear speech.) [3.2 seconds duration] Te esperaré al otro lado.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "516d39dfb8093ac01a4bc6413a0f0d3e2d7d6b70144e20b46b55a63e4ed0f669", "generation_seconds_batch": 23.485097950091586, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.633519411087036, "target_z": 1.4765625675499476, "ce": 5.7684478759765625, "genuineness": 4.6203155517578125, "identity": 0.3787904977798462, "hypothesis": "La carretera gira a la izquierda tras el puente. Sigue hasta la pequeña puerta de madera. Te esperaré al otro lado.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b6c5e532cb1fc50e3595a0ee36dcf6320970276d74dedd28a266768325221cc1.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "b6cf834d64aed7e9497ea291f14367cb163a27b720c81eaf820a0ec92eb84e14", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6b8b7ff450e43c77a26b3bdf0a5e8011191f5089a2a14bd210b016ff74812252", "generation_seconds_batch": 23.50608081498649, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0807387828826904, "target_z": -0.37499995676803355, "ce": 6.252951145172119, "genuineness": 4.394378662109375, "identity": 0.7626739740371704, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b6cf834d64aed7e9497ea291f14367cb163a27b720c81eaf820a0ec92eb84e14.mp3" }, { "id": "heldout-en-1", "language": "en", "sentences": [ "The keys are on the kitchen table.", "I checked the drawer... then looked beside the window.", "Take them with you when you leave." ], "text": "The keys are on the kitchen table. I checked the drawer... then looked beside the window. Take them with you when you leave.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-1", "key": "b6ece82e3c75c002216a5b2f3671b9c4b50ffe4901d71a42bc389e49c888cc5c", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The keys are on the kitchen table.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Take them with you when you leave.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The keys are on the kitchen table.\n(Natural clear speech.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Natural clear speech.) [3.2 seconds duration] Take them with you when you leave.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "358b936880cc8cc8f75403d56d21177236357b97a38230ffd6a7b5f6f0b2e6e5", "generation_seconds_batch": 16.069452674943022, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.159585475921631, "target_z": 1.1328125783579392, "ce": 5.919358730316162, "genuineness": 4.005263805389404, "identity": 0.3041524887084961, "hypothesis": "The keys are on the kitchen table. I checked the drawer, then looked beside the window. Take them with you when you", "wer": 0.043478260869565216, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b6ece82e3c75c002216a5b2f3671b9c4b50ffe4901d71a42bc389e49c888cc5c.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "b71273e1e637fd1342a80bc7528a19352b15cf48c749ade17ee2147970dbdea6", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "9c425cfee4542b0719e9add3e623590ff29eebc93619b61f84877d3e213b4f7d", "generation_seconds_batch": 22.30017531698104, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.13804292678833, "target_z": 1.1171875081059937, "ce": 6.0146708488464355, "genuineness": 4.331617832183838, "identity": 0.3193134367465973, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b71273e1e637fd1342a80bc7528a19352b15cf48c749ade17ee2147970dbdea6.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "b717b3a2004c4239825276a11597fa45d245db735c5a3aeaf3c23162e13dbaa4", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "47bc02fc386f0b9aaabb2e00a60230db5cf2a9867887ed0f4f5e5e19798f9217", "generation_seconds_batch": 22.31042186799459, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.72162663936615, "target_z": 0.089843721629022, "ce": 6.2767791748046875, "genuineness": 4.444586753845215, "identity": 0.6777796745300293, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b717b3a2004c4239825276a11597fa45d245db735c5a3aeaf3c23162e13dbaa4.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "b7272aae230bb642d82251d89c426d4e30844be30c01161f3113288a71655bee", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "afc01b05690bebbbc41120130c557ccd9e9d5b3561c425d9bc64e971d8999534", "generation_seconds_batch": 16.63694982707966, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0626444816589355, "target_z": 1.0625001080799161, "ce": 6.252951145172119, "genuineness": 4.193545341491699, "identity": 0.3310694694519043, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b7272aae230bb642d82251d89c426d4e30844be30c01161f3113288a71655bee.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "b738895fffed99300914f83a5ae9c982e7568a4711f3c0b3cc9caac68d62646e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e641b761f96bd935dfabb6e0a69d33ada0d61daaea3b4c63287fd51732e2f5dd", "generation_seconds_batch": 23.556757037062198, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.773545265197754, "target_z": 1.578125005403996, "ce": 5.649308204650879, "genuineness": 4.394378662109375, "identity": 0.13425135612487793, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b738895fffed99300914f83a5ae9c982e7568a4711f3c0b3cc9caac68d62646e.mp3" }, { "id": "heldout-de-3", "language": "de", "sentences": [ "Jemand hat den kleinen blauen Stuhl umgestellt.", "Er stand hier... neben dem hohen Bücherregal.", "Wir stellen ihn nach dem Essen zurück." ], "text": "Jemand hat den kleinen blauen Stuhl umgestellt. Er stand hier... neben dem hohen Bücherregal. Wir stellen ihn nach dem Essen zurück.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-3", "key": "b7b7f7b9ff86419d1b31f458bc2c4e2cb891223bbac371ddad105deb0658de6c", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Natural clear speech.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Natural clear speech.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "eee9c3b2bebe8e40197cea680298f824c35693b5eceefa05702f90eef4c63513", "generation_seconds_batch": 22.946463749045506, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9764745235443115, "target_z": 1.0, "ce": 5.911416053771973, "genuineness": 4.545003414154053, "identity": 0.45262694358825684, "hypothesis": "Jemand hat den kleinen blauen Stuhl umgestellt. Er stand hier, neben dem hohen Bücherregal. Wir stellen ihn nach dem Essen zurück.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b7b7f7b9ff86419d1b31f458bc2c4e2cb891223bbac371ddad105deb0658de6c.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "b7c94f02924235ef09ca55e6d6f50fac4610316723a6851438716d566d7f38da", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "41fae4241e563e8b64fc4e90d08517f41da3db837df70c7bbfd234164ac89afc", "generation_seconds_batch": 17.07711539103184, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.527531147003174, "target_z": 2.1250000432319665, "ce": 5.689021587371826, "genuineness": 4.582659721374512, "identity": 0.7302351593971252, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b7c94f02924235ef09ca55e6d6f50fac4610316723a6851438716d566d7f38da.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "b7d72fd31367571b1939c3d6c417f48402fef5123f45979c3b35aef73aeb38fb", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "2ce4ed77ead40e6f4b80b5e15245cfd5264a3a5bdea896d5aafd953fdfaf1c83", "generation_seconds_batch": 23.205556552973576, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.588710308074951, "target_z": 0.7187499459600419, "ce": 6.268836498260498, "genuineness": 4.507347106933594, "identity": 0.6022179126739502, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b7d72fd31367571b1939c3d6c417f48402fef5123f45979c3b35aef73aeb38fb.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "b80d36bbab7dcdfc54364c58f48deaea14264ed26abb799889de3cc770813149", "bucket": "vn_S_ASMR_reg", "phase": "wording", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a96107246b3acba4fd38b3c8946810b0856095243e2e5c8f1aeeeabd249ad14c", "generation_seconds_batch": 16.846011584973894, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.874147891998291, "target_z": 0.9257812621589906, "ce": 6.41180419921875, "genuineness": 4.557555198669434, "identity": 0.5250365734100342, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b80d36bbab7dcdfc54364c58f48deaea14264ed26abb799889de3cc770813149.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "b8868e12d315757a63b39e9cb25699e0e9074a5ce1e80f5c6a6a3b3d8e610c2c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3181e114fbd72218b899492c1b29619f9adbfb8b658514580fd9ffabcb5f795f", "generation_seconds_batch": 16.22605377691798, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.741231679916382, "target_z": 1.5546875729539433, "ce": 5.502368927001953, "genuineness": 4.419482707977295, "identity": 0.5268322229385376, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b8868e12d315757a63b39e9cb25699e0e9074a5ce1e80f5c6a6a3b3d8e610c2c.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "b88e461d9c090bcd5486abc4b952b9bdf601ab3117bf27a80828667e3374f700", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "91bfff6be990c090b604a94bad04d19d60d9ee5ea6c9446ed89d31b36115d264", "generation_seconds_batch": 22.24144284101203, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9674272537231445, "target_z": 1.718749945960042, "ce": 6.046441555023193, "genuineness": 4.570107460021973, "identity": 0.6914421319961548, "hypothesis": "La porte est encore ouverte, j'ai posé ton manteau sur la chaise, dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b88e461d9c090bcd5486abc4b952b9bdf601ab3117bf27a80828667e3374f700.mp3" }, { "id": "heldout-de-4", "language": "de", "sentences": [ "Ich schrieb die Nummer auf den Umschlag.", "Lies sie genau... und prüfe die letzten beiden Ziffern.", "Ruf an, wenn du Zeit hast." ], "text": "Ich schrieb die Nummer auf den Umschlag. Lies sie genau... und prüfe die letzten beiden Ziffern. Ruf an, wenn du Zeit hast.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-4", "key": "b8af64a496532d2375232bcb2fafa9d500b0c78afc0f106855cefc1b46e99b58", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Natural clear speech.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Natural clear speech.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a0eba770a3763687225645783d9d03af6539dffa4b767834a467c10b051df2f1", "generation_seconds_batch": 23.28906766604632, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.461179733276367, "target_z": 1.3515625243179812, "ce": 5.998785495758057, "genuineness": 4.582659721374512, "identity": 0.6350678205490112, "hypothesis": "Ich schrieb die Nummer auf den Umschlag, ließ sie genau und prüfe die letzten beiden Ziffern. Ruf an, wenn du Zeit hast.", "wer": 0.045454545454545456, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b8af64a496532d2375232bcb2fafa9d500b0c78afc0f106855cefc1b46e99b58.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "b9527c141e7f49e6fb0b477c8eb2b0a7070ef6bb74eea57e78cf9decb2cb7ef4", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d5ad3e95f2d93260aff22ec99ff81ebbdf703c5a82a50f198ff2cec0bd42c2e0", "generation_seconds_batch": 22.459264148026705, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.491769313812256, "target_z": 0.6484374756820188, "ce": 6.4515180587768555, "genuineness": 4.6705241203308105, "identity": 0.6544903516769409, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b9527c141e7f49e6fb0b477c8eb2b0a7070ef6bb74eea57e78cf9decb2cb7ef4.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "b958fd243f0e31b0050917e28e62e3e3dc0ffb3f03fce8134476111e3c1f930a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d3e715bf9773c32216b5974d95eabf6b4f8b238361fb6a69bf691f2fc34e89dc", "generation_seconds_batch": 25.695698769995943, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1488142013549805, "target_z": 1.1250000432319665, "ce": 6.197352409362793, "genuineness": 4.6203155517578125, "identity": 0.5413674116134644, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b958fd243f0e31b0050917e28e62e3e3dc0ffb3f03fce8134476111e3c1f930a.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "b971e6a3771143f452f47c5b22467fcb80690bc987655951dcb82f2fb65e07ae", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "72e584eb2a35ec8d65454e81cc0d3196f826111efff8fb863a67f0b3479ac712", "generation_seconds_batch": 17.32473435602151, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1272716522216797, "target_z": 1.1093749729800209, "ce": 6.252951145172119, "genuineness": 4.557555198669434, "identity": 0.3607524037361145, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b971e6a3771143f452f47c5b22467fcb80690bc987655951dcb82f2fb65e07ae.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "b9d4db41bbbdc7b0d8158b766cec90d836aaf4b45ac287008a6168b76daa1bda", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e3a6c44b351e490446a95ede2d4297613737d509157c60c4c6451bca6586d908", "generation_seconds_batch": 22.98898829310201, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8849191665649414, "target_z": 0.9335937972849634, "ce": 6.324435234069824, "genuineness": 4.419482707977295, "identity": 0.5595927238464355, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschalten.", "wer": 0.0625, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/b9d4db41bbbdc7b0d8158b766cec90d836aaf4b45ac287008a6168b76daa1bda.mp3" }, { "id": "heldout-fr-1", "language": "fr", "sentences": [ "Les clés sont sur la table de la cuisine.", "J’ai regardé dans le tiroir... puis près de la fenêtre.", "Prends-les quand tu partiras." ], "text": "Les clés sont sur la table de la cuisine. J’ai regardé dans le tiroir... puis près de la fenêtre. Prends-les quand tu partiras.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-1", "key": "ba70e7e208a6b15a9236ccd5a69e0e5fd403df2a937202d690effa2cdc45d9f2", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Prends-les quand tu partiras.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Prends-les quand tu partiras.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "65ce72dc643b3cf6b78b752ed0d05db0d9ca973547459ce7e0c1cdc9148710b4", "generation_seconds_batch": 16.290645889006555, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.752593994140625, "target_z": 0.11230471688422722, "ce": 6.316492557525635, "genuineness": 4.413206577301025, "identity": 0.7459250688552856, "hypothesis": "Les clés sont sur la table de la cuisine. J'ai regardé dans le tiroir, puis près de la fenêtre. Prends-les quand tu partiras.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ba70e7e208a6b15a9236ccd5a69e0e5fd403df2a937202d690effa2cdc45d9f2.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "bb3ec1845cf5f903778909aa37f23223c76cf44ecf1dbbd54ba756f7f9f0728a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "71c6600e4cf7b91d742108785b1c6eb416c03bea656b4f84828813253bae2780", "generation_seconds_batch": 17.582223370089196, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.042825698852539, "target_z": 1.7734373459861195, "ce": 5.411027908325195, "genuineness": 4.256305694580078, "identity": 0.3801383674144745, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/bb3ec1845cf5f903778909aa37f23223c76cf44ecf1dbbd54ba756f7f9f0728a.mp3" }, { "id": "heldout-de-2", "language": "de", "sentences": [ "Die Straße führt hinter der Brücke nach links.", "Geh weiter... bis zu dem kleinen Holztor.", "Ich warte auf der anderen Seite." ], "text": "Die Straße führt hinter der Brücke nach links. Geh weiter... bis zu dem kleinen Holztor. Ich warte auf der anderen Seite.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-2", "key": "bb5eef4854f22452c38c04f4ab7b18e8e2aeb71267927adef1fe4ab311eb4370", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Natural clear speech.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Natural clear speech.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "4bd2cde340321eb4ee080889a992fa2136d9406cb42241e4607a444c7e73c55e", "generation_seconds_batch": 15.984086344949901, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.687375545501709, "target_z": 1.5156250702519456, "ce": 5.919358730316162, "genuineness": 4.532451152801514, "identity": 0.44438302516937256, "hypothesis": "Die Straße führt hinter der Brücke nach links. Geh weiter bis zu dem kleinen Holztor. Ich warte auf der anderen Seite.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/bb5eef4854f22452c38c04f4ab7b18e8e2aeb71267927adef1fe4ab311eb4370.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "bbb316df3100e251ce2c1a757e9259318fb807f17cab7e40a0aadb4e4309894a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c53ec5fe163f5c0f27e361e883a595ee6e4906a06e5e5949e1e046a682baf1fa", "generation_seconds_batch": 25.090834756032564, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6112215518951416, "target_z": 0.009765582443532999, "ce": 6.125868320465088, "genuineness": 4.413206577301025, "identity": 0.5076266527175903, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/bbb316df3100e251ce2c1a757e9259318fb807f17cab7e40a0aadb4e4309894a.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "bbc9ac339e9ff5961495131bc1823be6ceef32c742f5daab0478b92a415b8ad4", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "bb0343fdd5423121bf3a4bc68093ce767284f0ff3ccf168b2836f01e0409862f", "generation_seconds_batch": 23.172279790975153, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.497154951095581, "target_z": 0.6523437432450052, "ce": 6.133810997009277, "genuineness": 4.457139015197754, "identity": 0.6638777852058411, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/bbc9ac339e9ff5961495131bc1823be6ceef32c742f5daab0478b92a415b8ad4.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "bbf9da9316ee4e9259a452372dd41269c7c366828922ac310c61dd2b8547efb2", "bucket": "vn_S_ASMR_reg", "phase": "wording", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f770f0f9fd79642cdd1047b73576fe465cb0deca0cae01d09a4113eb8012d203", "generation_seconds_batch": 16.683995866100304, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8441493511199951, "target_z": 0.17871091959926388, "ce": 6.403861999511719, "genuineness": 4.413206577301025, "identity": 0.2177809476852417, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/bbf9da9316ee4e9259a452372dd41269c7c366828922ac310c61dd2b8547efb2.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "bc75ef0076cd582960e4ae68b1b35464223a5789b6da59ad799a93ce30542db1", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "88f941b278fdd17df8916d6637d409c8638671f293609c622254a246ef93e9c6", "generation_seconds_batch": 24.60167984396685, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1184380054473877, "target_z": -0.3476562567549948, "ce": 6.260893821716309, "genuineness": 4.545003414154053, "identity": 0.6764246821403503, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/bc75ef0076cd582960e4ae68b1b35464223a5789b6da59ad799a93ce30542db1.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "bc7c3ed86bb9b29b545a6767c6599db17d0d48f4b01d312b14a92dc09f333ddd", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "10a12ad58b8b2bf3056cb80177003e73f990c54eb75879d189a588992fb5e7a0", "generation_seconds_batch": 22.31042186799459, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.68123459815979, "target_z": 0.06054688783449005, "ce": 6.221180438995361, "genuineness": 4.3504462242126465, "identity": 0.5478002429008484, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/bc7c3ed86bb9b29b545a6767c6599db17d0d48f4b01d312b14a92dc09f333ddd.mp3" }, { "id": "heldout-es-2", "language": "es", "sentences": [ "La carretera gira a la izquierda tras el puente.", "Sigue... hasta la pequeña puerta de madera.", "Te esperaré al otro lado." ], "text": "La carretera gira a la izquierda tras el puente. Sigue... hasta la pequeña puerta de madera. Te esperaré al otro lado.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-2", "key": "bca843f0a2588e8d57e1f3800cfd4bce661e708b346da9d7f098948f9337199f", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Te esperaré al otro lado.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Natural clear speech.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Natural clear speech.) [3.2 seconds duration] Te esperaré al otro lado.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "654f375e6430a197dd47b8bed4be0c0517093d7fc4335e444e78b1083fd3ee70", "generation_seconds_batch": 17.201339381048456, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3857812881469727, "target_z": 1.2968751242919037, "ce": 5.919358730316162, "genuineness": 4.557555198669434, "identity": 0.7368886470794678, "hypothesis": "La carretera gira a la izquierda tras el puente. Sigue hasta la pequeña puerta de madera. Te esperaré al otro lado.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/bca843f0a2588e8d57e1f3800cfd4bce661e708b346da9d7f098948f9337199f.mp3" }, { "id": "heldout-es-0", "language": "es", "sentences": [ "Tu mensaje llegó después de que me fuera.", "Lo leí dos veces... y guardé el teléfono.", "Dime cuándo podemos hablar." ], "text": "Tu mensaje llegó después de que me fuera. Lo leí dos veces... y guardé el teléfono. Dime cuándo podemos hablar.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-0", "key": "bcbedb3e0f4ea508c09338a2b703859116e030693c7d2cf7c993b261e3606b64", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo podemos hablar.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Natural clear speech.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo podemos hablar.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "34d1cd9727299a7e23f9180ed7977aabcc9931a6bfbb03d6c16a4f8558af07c8", "generation_seconds_batch": 23.700036367052235, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.461179733276367, "target_z": 1.3515625243179812, "ce": 5.242246150970459, "genuineness": 4.168440818786621, "identity": 0.366249144077301, "hypothesis": "Tu mensaje llegó después de que me fuera, lo leí dos veces y guardé el teléfono, dime cuándo podemos apuntar.", "wer": 0.05, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/bcbedb3e0f4ea508c09338a2b703859116e030693c7d2cf7c993b261e3606b64.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "bddcc3c667fdcfa84d5e1a28f378888604974909a9c0eb736e2c4fa5b769c778", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3a2b4ffcaa32a0a743979ca3548897a99a594abeae2175ca3903f8f12c09d00f", "generation_seconds_batch": 25.12463709106669, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.031299352645874, "target_z": 0.31445306893354347, "ce": 6.324435234069824, "genuineness": 4.6203155517578125, "identity": 0.6744158267974854, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/bddcc3c667fdcfa84d5e1a28f378888604974909a9c0eb736e2c4fa5b769c778.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "bf093ab2a2318d52ec9960749513f4762bd43e5a2bad4da02edf813ec8a77e39", "bucket": "vn_S_ASMR_reg", "phase": "wording", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "410c24a21115653092d932708d2b1bbf936614504cbe0a390301f120015c6c0c", "generation_seconds_batch": 16.683995866100304, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9909073114395142, "target_z": 0.2851562351390115, "ce": 6.3641486167907715, "genuineness": 4.6203155517578125, "identity": 0.33686143159866333, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/bf093ab2a2318d52ec9960749513f4762bd43e5a2bad4da02edf813ec8a77e39.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "bf2ea67e6ff58db80b3a9801b0ab4bba559d1234f39b963b5f49e3209b00ce49", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "69c42aba03872380d0c5d5e8525b379500210d71a9a278bb9d287eab0890b44e", "generation_seconds_batch": 22.712467964971438, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.364238739013672, "target_z": 1.281250054039958, "ce": 5.474569320678711, "genuineness": 4.168440818786621, "identity": 0.5023444294929504, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in klaar.", "wer": 0.05555555555555555, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/bf2ea67e6ff58db80b3a9801b0ab4bba559d1234f39b963b5f49e3209b00ce49.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "bf33f40c27aa51b819bd70b05a3b1746a3b0c800a20187ea3cf206e41ebf8273", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "56152a2327caa1440a8eeaacee62532c4a9cf90b4f47f3074d46b74c9ae0fe6d", "generation_seconds_batch": 23.60011624090839, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.150538444519043, "target_z": 1.851562697245847, "ce": 5.911416053771973, "genuineness": 4.6203155517578125, "identity": 0.7619005441665649, "hypothesis": "La porte est encore ouverte, j'ai posé ton manteau sur la chaise, dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/bf33f40c27aa51b819bd70b05a3b1746a3b0c800a20187ea3cf206e41ebf8273.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "c0a11a19205132d423f54ef02fd32a2339244845a46ca390d32db685cac0b0b0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "095f91550837dbaa986c3a7373f83ad76e65e6165ce6b8619b5feee59f4800f7", "generation_seconds_batch": 18.15410049201455, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2349839210510254, "target_z": 1.1874999783840168, "ce": 5.847874641418457, "genuineness": 4.331617832183838, "identity": 0.6302688121795654, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c0a11a19205132d423f54ef02fd32a2339244845a46ca390d32db685cac0b0b0.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "c0fba0f8f85544d3a8eadd84fddbd0195a3119dbbd062b060ccd4a06ec4646f3", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "6e135b2cb3ae2fb5af4f737fed8ebe38e9fc9dfb5408605afda603263df8bff5", "generation_seconds_batch": 22.28326332499273, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1811277866363525, "target_z": 1.1484374756820188, "ce": 6.205295085906982, "genuineness": 4.381826400756836, "identity": 0.7239817976951599, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c0fba0f8f85544d3a8eadd84fddbd0195a3119dbbd062b060ccd4a06ec4646f3.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "c10b5b2b6f7d5c2a70b92d85d801aac0498995e5a0738f83ceeea05db3b1650d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "3436c8a49325362ed06a03917701dfcb1df40aef6b718de159e815e7dfa07d92", "generation_seconds_batch": 17.474135274998844, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.396552324295044, "target_z": 1.3046874864900104, "ce": 5.689021587371826, "genuineness": 4.268857479095459, "identity": 0.4515850841999054, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c10b5b2b6f7d5c2a70b92d85d801aac0498995e5a0738f83ceeea05db3b1650d.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "c12dd86e51420a08367793bcf52b2649503286f77686f8a6c234f272edc7cbf2", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ee1b38f88528d5cd32d921a817da06a4afdd140e159edf1c804d310e078de8ee", "generation_seconds_batch": 23.403526039095595, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0716915130615234, "target_z": 0.3437499891920084, "ce": 6.395919322967529, "genuineness": 4.770940780639648, "identity": 0.5557109117507935, "hypothesis": "I found the little box. It was beside the old window. We can look inside together,", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c12dd86e51420a08367793bcf52b2649503286f77686f8a6c234f272edc7cbf2.mp3" }, { "id": "heldout-nl-0", "language": "nl", "sentences": [ "Je bericht kwam nadat ik vertrokken was.", "Ik las het twee keer... en legde mijn telefoon weg.", "Zeg maar wanneer we kunnen praten." ], "text": "Je bericht kwam nadat ik vertrokken was. Ik las het twee keer... en legde mijn telefoon weg. Zeg maar wanneer we kunnen praten.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-0", "key": "c1689af0ad55ccb158ad337b9bdd668f28b6241874f684682f0f24fbeddbe673", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Natural clear speech.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "12394adc505fea4de7cbce6b017e7d48210de9a495aba966c7b54e23e731b907", "generation_seconds_batch": 25.183655660948716, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.051873207092285, "target_z": 1.0546875729539433, "ce": 5.458683967590332, "genuineness": 4.105680465698242, "identity": 0.35444188117980957, "hypothesis": "Je bericht kwam nadat ik vertrokken was. Ik las ze twee keer en legde mijn telefoon weg. Zeg maar wanneer we kunnen praten.", "wer": 0.043478260869565216, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c1689af0ad55ccb158ad337b9bdd668f28b6241874f684682f0f24fbeddbe673.mp3" }, { "id": "heldout-en-2", "language": "en", "sentences": [ "The road turns left behind the bridge.", "Keep going... until you reach the small wooden gate.", "I will meet you on the other side." ], "text": "The road turns left behind the bridge. Keep going... until you reach the small wooden gate. I will meet you on the other side.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-2", "key": "c192a290eb659c96c6725355024bd4ba79459a7e6e6be8dd47a785afa86c7aa3", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The road turns left behind the bridge.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I will meet you on the other side.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The road turns left behind the bridge.\n(Natural clear speech.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Natural clear speech.) [3.2 seconds duration] I will meet you on the other side.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5c0ae76f68ac453f3d1296382134b68a6db6d70444a1a39de83c5ac2c48aa399", "generation_seconds_batch": 25.159405322978273, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8903045654296875, "target_z": 0.9374998919200838, "ce": 5.657250881195068, "genuineness": 4.068024158477783, "identity": 0.480263888835907, "hypothesis": "The road turns left behind the bridge. Keep going until you reach the small wooden gate. I will meet you on the other side.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c192a290eb659c96c6725355024bd4ba79459a7e6e6be8dd47a785afa86c7aa3.mp3" }, { "id": "heldout-fr-0", "language": "fr", "sentences": [ "Ton message est arrivé après mon départ.", "Je l’ai relu... puis j’ai rangé mon téléphone.", "Dis-moi quand nous pourrons parler." ], "text": "Ton message est arrivé après mon départ. Je l’ai relu... puis j’ai rangé mon téléphone. Dis-moi quand nous pourrons parler.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-0", "key": "c2b3dee87d72071075cdbeb79d8afae20d64749fb32254cab41e305a03c89f53", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Natural clear speech.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "48ed051ad39b016bf6b2aeb050ea011d0488bca9960903b84ce064ab0c54a6af", "generation_seconds_batch": 18.42447387694847, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.085910797119141, "target_z": 1.8046874864900104, "ce": 5.847874641418457, "genuineness": 4.68307638168335, "identity": 0.9193089604377747, "hypothesis": "Ton message est arrivé après mon départ, je l'ai relu, puis j'ai rangé mon téléphone. Dis-moi quand nous pourrons parler.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c2b3dee87d72071075cdbeb79d8afae20d64749fb32254cab41e305a03c89f53.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "c2b61c515acd9f31bb464336a61a00e7dca5667f4d20fd755c6e62943d47f753", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "80d1e3ba721bfe0cf044e21acb1131f29f6f478385413774d6761e0e38a53441", "generation_seconds_batch": 24.523894718964584, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.279792785644531, "target_z": 1.9453124270460567, "ce": 5.673136234283447, "genuineness": 4.482243061065674, "identity": 0.6348544359207153, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce calcul.", "wer": 0.10526315789473684, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c2b61c515acd9f31bb464336a61a00e7dca5667f4d20fd755c6e62943d47f753.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "c2e7a9904f97b2dea8366af961f29d6b1119b20ff2482f913fb92e677201d539", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "93d526bc92d000d5b6aaabd0d1e57d0c651ff76b9fb71f8ab39e5be5957ea6b2", "generation_seconds_batch": 22.414825026062317, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.139766693115234, "target_z": 1.8437498162641426, "ce": 5.788304805755615, "genuineness": 4.582659721374512, "identity": 0.7173939943313599, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c2e7a9904f97b2dea8366af961f29d6b1119b20ff2482f913fb92e677201d539.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "c2e99dabf7cd81a8cc7f25ddddb750bc3e01830e9a1bb9ad3ceeb4d2b6592cc6", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "43b57066991ed5cb6d7923f114e7749c4e7a8481abb2cfc313fc6ea580914aed", "generation_seconds_batch": 23.610983759048395, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2996113300323486, "target_z": 1.2343750162119875, "ce": 6.229123115539551, "genuineness": 4.193545341491699, "identity": 0.46910884976387024, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c2e99dabf7cd81a8cc7f25ddddb750bc3e01830e9a1bb9ad3ceeb4d2b6592cc6.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "c2f30249846d52c9c1bba76dcd54aad1762da26d7d8ac02bfa674c365b8bf4c1", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "16d67825ad134dc51b5343e7bb917250973744780319f907afbf2d2570c4132e", "generation_seconds_batch": 23.898921641986817, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2342287302017212, "target_z": -0.2636718500065194, "ce": 6.316492557525635, "genuineness": 4.068024158477783, "identity": 0.48806053400039673, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c2f30249846d52c9c1bba76dcd54aad1762da26d7d8ac02bfa674c365b8bf4c1.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "c2ff189cf29c60a494bc24927476edf4577edab9abd6f7259e77bad65d6a6aa8", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "51a618f111f97b75823bbc4ef5de2a2ababc2afd72595f936217a1e13077932d", "generation_seconds_batch": 24.21342054498382, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8643454313278198, "target_z": 0.19335937972849634, "ce": 6.419747352600098, "genuineness": 4.344170093536377, "identity": 0.6104546189308167, "hypothesis": "Ich habe die kleine Schacht gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0625, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c2ff189cf29c60a494bc24927476edf4577edab9abd6f7259e77bad65d6a6aa8.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "c343b4af051444b87ce4a382c63d1719be284a34701a209e64c28d9d7742cb8e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "87f0833107c99662f35e453cc8f2f5be5ac58bfa16448fe248ea33f82624f5d9", "generation_seconds_batch": 17.800323338015005, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.333648681640625, "target_z": 1.9843747568201886, "ce": 6.022613525390625, "genuineness": 4.7834930419921875, "identity": 0.7794883251190186, "hypothesis": "J'ai trouvé la petite boîte, elle était près de la vieille fenêtre, nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c343b4af051444b87ce4a382c63d1719be284a34701a209e64c28d9d7742cb8e.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "c36b4d900aa4e8865158bea4df1a7ea1c0ec51b6080c405f6840f116af0ef962", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "c0e255192a1192b28af8d38f55b10134533741fc5e2f7d2531e7f316223e0710", "generation_seconds_batch": 23.03462289494928, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.655061721801758, "target_z": 1.4921874648740272, "ce": 5.982900142669678, "genuineness": 4.432034492492676, "identity": 0.29758816957473755, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c36b4d900aa4e8865158bea4df1a7ea1c0ec51b6080c405f6840f116af0ef962.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "c3716582a8e72f55ae66bb84984237b1a253f62de17457d27ce304f11d848057", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "aed4333b6c15597ff778da3ccc0ae705b7ba6c00c84d677422b197b6ad7694ae", "generation_seconds_batch": 23.668342157965526, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.51503586769104, "target_z": 1.3906250270199791, "ce": 6.229123115539551, "genuineness": 4.7332844734191895, "identity": 0.869260847568512, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c3716582a8e72f55ae66bb84984237b1a253f62de17457d27ce304f11d848057.mp3" }, { "id": "heldout-es-2", "language": "es", "sentences": [ "La carretera gira a la izquierda tras el puente.", "Sigue... hasta la pequeña puerta de madera.", "Te esperaré al otro lado." ], "text": "La carretera gira a la izquierda tras el puente. Sigue... hasta la pequeña puerta de madera. Te esperaré al otro lado.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-2", "key": "c3977db0634952a7f9e4083e6135fbbc87199d34c7cb8c3338515d9ae38c18f4", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Te esperaré al otro lado.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Te esperaré al otro lado.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "77d2925d0c024c33a17d828945a974754392d35caedb3e66b1a6e353d4ff1acb", "generation_seconds_batch": 17.443841414991766, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9478224515914917, "target_z": 0.2539062675629864, "ce": 6.0305562019348145, "genuineness": 4.607763767242432, "identity": 0.6447604894638062, "hypothesis": "La carretera gira a la izquierda tras el puente. Sigue hasta la pequeña puerta de madera. Te esperaré al otro lado.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c3977db0634952a7f9e4083e6135fbbc87199d34c7cb8c3338515d9ae38c18f4.mp3" }, { "id": "heldout-de-0", "language": "de", "sentences": [ "Deine Nachricht kam, als ich schon weg war.", "Ich las sie zweimal... dann legte ich das Handy weg.", "Sag mir, wann wir reden können." ], "text": "Deine Nachricht kam, als ich schon weg war. Ich las sie zweimal... dann legte ich das Handy weg. Sag mir, wann wir reden können.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-0", "key": "c39a0e8c55b8fb3a776a1389d63d1c0c55460b64e23041dd941632a0efcc42c0", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wann wir reden können.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Natural clear speech.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wann wir reden können.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "666deb4e33d58a242f0dc8c922c7950e67a21638c912c621c61a0ad071d30cd0", "generation_seconds_batch": 17.012787114013918, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.601205587387085, "target_z": 1.4531249621720292, "ce": 5.7684478759765625, "genuineness": 4.519899368286133, "identity": 0.5577229857444763, "hypothesis": "Deine Nachricht kam, als ich schon weg war. Ich las sie zweimal, dann legte ich das Handy weg. Sag mir, wann wir reden können.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c39a0e8c55b8fb3a776a1389d63d1c0c55460b64e23041dd941632a0efcc42c0.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "c3b8f4b9a5ca40d090bedf83f6269f5fb8aeb9fefafc1517a4f9644ab0efaa9b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "a8367a5e218ebbbc8a9852f453739615d126a6b8f7fa1f9e486cd37d4086190d", "generation_seconds_batch": 25.54708585201297, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.118224143981934, "target_z": 1.828124746012197, "ce": 6.038498878479004, "genuineness": 4.7834930419921875, "identity": 0.7006812691688538, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c3b8f4b9a5ca40d090bedf83f6269f5fb8aeb9fefafc1517a4f9644ab0efaa9b.mp3" }, { "id": "heldout-nl-3", "language": "nl", "sentences": [ "Iemand heeft het kleine blauwe stoeltje verplaatst.", "Het stond hier... naast de hoge boekenkast.", "We zetten het na de lunch terug." ], "text": "Iemand heeft het kleine blauwe stoeltje verplaatst. Het stond hier... naast de hoge boekenkast. We zetten het na de lunch terug.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-3", "key": "c493619abd8899c9b3ce96f2a00a6bcdd900e874ef21436707b986365a8437ee", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We zetten het na de lunch terug.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We zetten het na de lunch terug.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f408ccdfcfd9f54b2ae50c9fa8ce377463e1a4bb98f84842549c244d6cefaa22", "generation_seconds_batch": 17.332933452911675, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1130523681640625, "target_z": -0.35156252431798113, "ce": 6.41180419921875, "genuineness": 4.65797233581543, "identity": 0.6403048038482666, "hypothesis": "Iemand heeft het kleine blauwe stoeltje verplaatst. Het stond hier. Naast de hoge boekenkast. We zetten het na de lunch terug.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c493619abd8899c9b3ce96f2a00a6bcdd900e874ef21436707b986365a8437ee.mp3" }, { "id": "heldout-es-4", "language": "es", "sentences": [ "Escribí el número en el sobre.", "Léelo bien... y comprueba las dos últimas cifras.", "Llámame cuando tengas un momento." ], "text": "Escribí el número en el sobre. Léelo bien... y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-4", "key": "c52c8e99ac851fdca320d3c471fbc6d840d296834f954996b2677895f85ce01c", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Escribí el número en el sobre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Llámame cuando tengas un momento.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Escribí el número en el sobre.\n(Natural clear speech.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Natural clear speech.) [3.2 seconds duration] Llámame cuando tengas un momento.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ce076488fad70757fc4de3604f8cd18ab45f17fff426b4d77b7976dc2d7708f9", "generation_seconds_batch": 16.72315739293117, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.105729103088379, "target_z": 1.0937499027280755, "ce": 5.816103935241699, "genuineness": 4.218649387359619, "identity": 0.4881654977798462, "hypothesis": "Escribe el número en el sobre, léelo bien y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "wer": 0.05263157894736842, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c52c8e99ac851fdca320d3c471fbc6d840d296834f954996b2677895f85ce01c.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "c52cc2775cea5b277abadd722405fef11c0b3f3a61d5f048f4a0da7c1f4bd92d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "925078042bbac6e7d1e68021b180821486ae6a458fbefb9dd88686db5f922d4f", "generation_seconds_batch": 23.205556552973576, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.033992290496826, "target_z": 0.3164062891789696, "ce": 6.395919322967529, "genuineness": 4.281409740447998, "identity": 0.6452787518501282, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c52cc2775cea5b277abadd722405fef11c0b3f3a61d5f048f4a0da7c1f4bd92d.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "c551075029fbb535c9d9efaf2e19680f514515c010e01409fa37c83ea1739cf7", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9114d9f35aa8abbfdfe41c38ddd6e9520e52eab48779ee71589a3d01f501f266", "generation_seconds_batch": 24.634708809899166, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8812575340270996, "target_z": 1.6562500108079916, "ce": 6.181467056274414, "genuineness": 4.268857479095459, "identity": 0.44357481598854065, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c551075029fbb535c9d9efaf2e19680f514515c010e01409fa37c83ea1739cf7.mp3" }, { "id": "heldout-fr-1", "language": "fr", "sentences": [ "Les clés sont sur la table de la cuisine.", "J’ai regardé dans le tiroir... puis près de la fenêtre.", "Prends-les quand tu partiras." ], "text": "Les clés sont sur la table de la cuisine. J’ai regardé dans le tiroir... puis près de la fenêtre. Prends-les quand tu partiras.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-1", "key": "c58a695380ebc5af0d64142b6ea004c8496790f123d0797714e3fc7ba2d48b05", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Prends-les quand tu partiras.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Natural clear speech.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Prends-les quand tu partiras.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1040db4d52d7eb5cb6222ed520d541647c5b86b0bce31bdc75ad60e675695484", "generation_seconds_batch": 22.506382621009834, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.06436824798584, "target_z": 1.789062416238065, "ce": 5.7128496170043945, "genuineness": 4.519899368286133, "identity": 0.5828639268875122, "hypothesis": "Les clés sont sur la table de la cuisine. J'ai regardé dans le tiroir, puis près de la fenêtre. Prends-les quand tu partiras.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c58a695380ebc5af0d64142b6ea004c8496790f123d0797714e3fc7ba2d48b05.mp3" }, { "id": "heldout-es-4", "language": "es", "sentences": [ "Escribí el número en el sobre.", "Léelo bien... y comprueba las dos últimas cifras.", "Llámame cuando tengas un momento." ], "text": "Escribí el número en el sobre. Léelo bien... y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-4", "key": "c5a9e509c7b3018659a3250dffd6f0b0f375d5fff24e76aedd54486d94f81ab3", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Escribí el número en el sobre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Llámame cuando tengas un momento.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Escribí el número en el sobre.\n(Natural clear speech.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Natural clear speech.) [3.2 seconds duration] Llámame cuando tengas un momento.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4100b7f49e3e03a23a9fe57a6d45f8de9874e1d4b3d1eb38aa59d964baf6a4bb", "generation_seconds_batch": 24.3650322050089, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7089178562164307, "target_z": 1.5312499675760252, "ce": 5.649308204650879, "genuineness": 4.432034492492676, "identity": 0.6146917939186096, "hypothesis": "Escribí el número en el sobre. Léelo bien y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c5a9e509c7b3018659a3250dffd6f0b0f375d5fff24e76aedd54486d94f81ab3.mp3" }, { "id": "heldout-en-3", "language": "en", "sentences": [ "Someone moved the little blue chair.", "It used to stand here... beside the tall bookshelf.", "We can put it back after lunch." ], "text": "Someone moved the little blue chair. It used to stand here... beside the tall bookshelf. We can put it back after lunch.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-3", "key": "c6bc8bdab84822beb2a67ab4c99a64916ae1593e62682a357af4fd97d42c4693", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Someone moved the little blue chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can put it back after lunch.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Someone moved the little blue chair.\n(Natural clear speech.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Natural clear speech.) [3.2 seconds duration] We can put it back after lunch.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "782587ba471324b9942616caa53dd210680425c5bf63efabc9b8b58055599d67", "generation_seconds_batch": 17.178146843914874, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.933389663696289, "target_z": 0.9687500324239748, "ce": 6.284721851348877, "genuineness": 4.356721878051758, "identity": 0.4387452006340027, "hypothesis": "Someone moved the little blue chair. It used to stand here beside the tall bookshelf. We can put it back after lunch.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c6bc8bdab84822beb2a67ab4c99a64916ae1593e62682a357af4fd97d42c4693.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "c765dac740a5f4b5ff175551f73746c47494375fd742fe8edb979c70c924f3b6", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "e48420163fc686eba8db4d757c547e4dd4ed0927ae4e8d6247c627459e9aaa18", "generation_seconds_batch": 21.895973603939638, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8812575340270996, "target_z": 1.6562500108079916, "ce": 5.391171455383301, "genuineness": 4.168440818786621, "identity": 0.5368645191192627, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás libre.", "wer": 0.06666666666666667, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c765dac740a5f4b5ff175551f73746c47494375fd742fe8edb979c70c924f3b6.mp3" }, { "id": "heldout-nl-0", "language": "nl", "sentences": [ "Je bericht kwam nadat ik vertrokken was.", "Ik las het twee keer... en legde mijn telefoon weg.", "Zeg maar wanneer we kunnen praten." ], "text": "Je bericht kwam nadat ik vertrokken was. Ik las het twee keer... en legde mijn telefoon weg. Zeg maar wanneer we kunnen praten.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-0", "key": "c82131bfaf89850d394bf3a21117deb94436b65306054c3e1cbac0c65e2f913d", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c43982dacf031a5d9a7fc97bc2137fff975ae2d42d826b68a0ba9a653fef7188", "generation_seconds_batch": 17.577438231906854, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8168436884880066, "target_z": -0.5664062459470032, "ce": 6.268836498260498, "genuineness": 4.519899368286133, "identity": 0.7271277904510498, "hypothesis": "Je bericht kwam nadat ik vertrokken was. Ik las het twee keer en legde mijn telefoon weg. Zeg maar wanneer we kunnen praten.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c82131bfaf89850d394bf3a21117deb94436b65306054c3e1cbac0c65e2f913d.mp3" }, { "id": "heldout-en-3", "language": "en", "sentences": [ "Someone moved the little blue chair.", "It used to stand here... beside the tall bookshelf.", "We can put it back after lunch." ], "text": "Someone moved the little blue chair. It used to stand here... beside the tall bookshelf. We can put it back after lunch.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-3", "key": "c8907269f81c0a33f09ba3b9d9f76b6acb2a8250e6881dbc11786cea01719e9d", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Someone moved the little blue chair.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can put it back after lunch.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Someone moved the little blue chair.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can put it back after lunch.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d7c6ed12fdbd2d35be258e24db66ef52ac578121272d9829cc617c92e7edb8e8", "generation_seconds_batch": 20.027848617057316, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7929860353469849, "target_z": 0.14160155067875915, "ce": 6.332377910614014, "genuineness": 4.494794845581055, "identity": 0.30305734276771545, "hypothesis": "Someone moved the little blue chair. It used to stand here beside the tall bookshelf. We can put it back after lunch.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c8907269f81c0a33f09ba3b9d9f76b6acb2a8250e6881dbc11786cea01719e9d.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "c8d48601e1e6b3a4544bc4c655c608a6ac95e182de933048cf3fa1cde26d954a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "355e57c8f37d87bd64f96547ee95f73d38cd8496adbe651f6c028abdf83915b2", "generation_seconds_batch": 31.204979359987192, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3948283195495605, "target_z": 0.5781250054039958, "ce": 6.427689552307129, "genuineness": 4.6203155517578125, "identity": 0.5192643404006958, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c8d48601e1e6b3a4544bc4c655c608a6ac95e182de933048cf3fa1cde26d954a.mp3" }, { "id": "heldout-de-0", "language": "de", "sentences": [ "Deine Nachricht kam, als ich schon weg war.", "Ich las sie zweimal... dann legte ich das Handy weg.", "Sag mir, wann wir reden können." ], "text": "Deine Nachricht kam, als ich schon weg war. Ich las sie zweimal... dann legte ich das Handy weg. Sag mir, wann wir reden können.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-0", "key": "c90d130d7351f3cd0a965b6503190727790fa7025d2cef88d6b9388d36d6b9b9", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wann wir reden können.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Natural clear speech.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wann wir reden können.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "b87067a610e39ad85eceba1f05f52360a10177a9853d598f954c8f4aeb9209ca", "generation_seconds_batch": 23.707801511045545, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2996113300323486, "target_z": 1.2343750162119875, "ce": 5.597680568695068, "genuineness": 4.532451152801514, "identity": 0.6098780632019043, "hypothesis": "Deine Nachricht kam, als ich schon weg war. Ich las sie zweimal. Dann legte ich das Handy weg. Sag mir, wann wir reden können.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c90d130d7351f3cd0a965b6503190727790fa7025d2cef88d6b9388d36d6b9b9.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "c97a5593582b4310e5c0052005cae410bd0ab9fe6e59d094779248a34c389818", "bucket": "vn_S_ASMR_reg", "phase": "wording", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "8502f0229f923e144ec197b142f4adfe18e9d875bbb5210a99a51da0177c6098", "generation_seconds_batch": 16.846011584973894, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.868384599685669, "target_z": 0.19628903716876964, "ce": 6.181467056274414, "genuineness": 4.243753433227539, "identity": 0.34316128492355347, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/c97a5593582b4310e5c0052005cae410bd0ab9fe6e59d094779248a34c389818.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "ca21daaa4aa92346ab81f80d9deef41b73d83cfa1d3d91d45b476c2de6b91d06", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1eb9f02ab5a4cce75c12cec9be8686e7a4178702d7b814cd689a0ecbe2121fae", "generation_seconds_batch": 15.981465880991891, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.010512351989746, "target_z": 1.750000086463933, "ce": 5.879645347595215, "genuineness": 4.494794845581055, "identity": 0.24798715114593506, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ca21daaa4aa92346ab81f80d9deef41b73d83cfa1d3d91d45b476c2de6b91d06.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "ca2fa22e22ef25de07856c90fdcffabea301c4c30fb21f46bbf9c5ced703ad40", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "a9392ebe14980bb2cdcd86d1d786770f33760ced0c0f1bc21fe08b458aaa3029", "generation_seconds_batch": 23.119388901046477, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4271419048309326, "target_z": 0.6015624378540482, "ce": 6.173524379730225, "genuineness": 4.632867813110352, "identity": 0.7541177272796631, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.10526315789473684, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ca2fa22e22ef25de07856c90fdcffabea301c4c30fb21f46bbf9c5ced703ad40.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "cac94b415879fedb5d5202290cd892312adb6e47e54e266fb3cf3ef642669442", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5c897607954c93ba0ef1d2ce6f748feca319aa7bdcbaa150e115c64d214c7410", "generation_seconds_batch": 24.00689779908862, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.750278949737549, "target_z": 0.8359376269939015, "ce": 6.38797664642334, "genuineness": 4.444586753845215, "identity": 0.3675634562969208, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/cac94b415879fedb5d5202290cd892312adb6e47e54e266fb3cf3ef642669442.mp3" }, { "id": "heldout-en-1", "language": "en", "sentences": [ "The keys are on the kitchen table.", "I checked the drawer... then looked beside the window.", "Take them with you when you leave." ], "text": "The keys are on the kitchen table. I checked the drawer... then looked beside the window. Take them with you when you leave.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-1", "key": "cad3115336b9ec75788982cc7ba8c181abb3890136c051d3e0f5fb4fd0103455", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The keys are on the kitchen table.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Take them with you when you leave.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The keys are on the kitchen table.\n(Natural clear speech.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Natural clear speech.) [3.2 seconds duration] Take them with you when you leave.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "106eaf21a9658639d442fd2f77d264cd3d756a60756ea1ad6ca2e2aa59b3592a", "generation_seconds_batch": 23.35967026592698, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.928004026412964, "target_z": 0.9648437648609884, "ce": 5.784333229064941, "genuineness": 3.9801597595214844, "identity": 0.45631033182144165, "hypothesis": "The keys are on the kitchen table. I checked the drawer, then looked beside the window. Take them with you when you like.", "wer": 0.043478260869565216, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/cad3115336b9ec75788982cc7ba8c181abb3890136c051d3e0f5fb4fd0103455.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "cae5c2b146f8812f31ac4a6d3be0ecd83e1f61f683e5f0f0b40ffcfefa36a25b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b382f3689c15c27caa82d499c84e77ff7b38bdd8223691ad7a23e87dc8d24740", "generation_seconds_batch": 24.121383397025056, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.637180805206299, "target_z": 0.7539061810990534, "ce": 6.125868320465088, "genuineness": 4.413206577301025, "identity": 0.3822503685951233, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/cae5c2b146f8812f31ac4a6d3be0ecd83e1f61f683e5f0f0b40ffcfefa36a25b.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "cb392c848db04f9a466ada87fee4a304221b9e74bf258fcb78c75c0fed06be33", "bucket": "vn_S_ASMR_reg", "phase": "wording", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b0febf89c2262f2c91ac40a7bfa065ee975c3328235b25df18bcc627704d4c52", "generation_seconds_batch": 18.048736938042566, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7718212604522705, "target_z": 0.8515625243179812, "ce": 6.316492557525635, "genuineness": 4.883909225463867, "identity": 0.7321274876594543, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/cb392c848db04f9a466ada87fee4a304221b9e74bf258fcb78c75c0fed06be33.mp3" }, { "id": "heldout-de-1", "language": "de", "sentences": [ "Die Schlüssel liegen auf dem Küchentisch.", "Ich suchte in der Schublade... dann neben dem Fenster.", "Nimm sie mit, wenn du gehst." ], "text": "Die Schlüssel liegen auf dem Küchentisch. Ich suchte in der Schublade... dann neben dem Fenster. Nimm sie mit, wenn du gehst.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-1", "key": "cb8d546b0514eab06330db09a998dc7da59a72ed85268fb897c3851d1d1b7089", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Natural clear speech.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Natural clear speech.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "19090295124f117cb721a527748de09107f6bc6e46aa4d31d14d21a47cdf98db", "generation_seconds_batch": 23.010773499961942, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.021283149719238, "target_z": 1.7578122757341739, "ce": 5.736677646636963, "genuineness": 4.432034492492676, "identity": 0.458662748336792, "hypothesis": "Die Schlüssel liegen auf dem Küchentisch. Ich suchte in der Schublade daneben dem Fenster. Nimm sie mit, wenn du gehst.", "wer": 0.09523809523809523, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/cb8d546b0514eab06330db09a998dc7da59a72ed85268fb897c3851d1d1b7089.mp3" }, { "id": "heldout-en-0", "language": "en", "sentences": [ "Your message arrived after I had left.", "I read it twice... before putting the phone away.", "Let me know when we can talk." ], "text": "Your message arrived after I had left. I read it twice... before putting the phone away. Let me know when we can talk.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-0", "key": "ccd1d0f7b71878dc3d264b033e9e17efc3def8c7800a08476394df0846d249c0", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Your message arrived after I had left.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Let me know when we can talk.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Your message arrived after I had left.\n(Natural clear speech.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Natural clear speech.) [3.2 seconds duration] Let me know when we can talk.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3166296187b2536e5f714cd872384721810243a0a9c39ba4f8346fa3190036b2", "generation_seconds_batch": 16.887940023094416, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.202670097351074, "target_z": 1.1640623730060984, "ce": 5.736677646636963, "genuineness": 4.369274139404297, "identity": 0.4041043519973755, "hypothesis": "Your message arrived after I had left. I read it twice, before putting the phone away. Let me know when we can talk.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ccd1d0f7b71878dc3d264b033e9e17efc3def8c7800a08476394df0846d249c0.mp3" }, { "id": "heldout-fr-2", "language": "fr", "sentences": [ "La route tourne à gauche après le pont.", "Continue... jusqu’à la petite barrière en bois.", "Je t’attendrai de l’autre côté." ], "text": "La route tourne à gauche après le pont. Continue... jusqu’à la petite barrière en bois. Je t’attendrai de l’autre côté.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-2", "key": "cd113f82fe1b2807b384263d27aca0d6c9f5f538089199c72459bddce1e105a4", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Natural clear speech.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Natural clear speech.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9d5891a0445caa60d29c7415fb7b30164a9876433d6bd77ba0ae8d54b8d1ffa1", "generation_seconds_batch": 15.536081769037992, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8812575340270996, "target_z": 1.6562500108079916, "ce": 5.77639102935791, "genuineness": 4.507347106933594, "identity": 0.8001418709754944, "hypothesis": "La route tourne à gauche après le pont, continue jusqu'à la petite barrière en bois. Je t'attendrai de l'autre côté.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/cd113f82fe1b2807b384263d27aca0d6c9f5f538089199c72459bddce1e105a4.mp3" }, { "id": "heldout-en-4", "language": "en", "sentences": [ "I wrote the number on the envelope.", "Read it carefully... and check the last two digits.", "Call me when you have a moment." ], "text": "I wrote the number on the envelope. Read it carefully... and check the last two digits. Call me when you have a moment.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-4", "key": "cd3ce1aab897db1e32bd930f37028b07fe08103268a1bc011757e82fe4ea5212", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I wrote the number on the envelope.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Call me when you have a moment.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I wrote the number on the envelope.\n(Natural clear speech.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Natural clear speech.) [3.2 seconds duration] Call me when you have a moment.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "8245f0817d4fab93c1303336f427b4c4304016cfda10a1ddcc0fd28405fd61c9", "generation_seconds_batch": 22.129737686016597, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6802659034729004, "target_z": 0.7851563216029445, "ce": 6.070269584655762, "genuineness": 3.9048471450805664, "identity": 0.4016871452331543, "hypothesis": "I wrote the number on the envelope. Read it carefully and check the last two digits. Call me when you have a moment.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/cd3ce1aab897db1e32bd930f37028b07fe08103268a1bc011757e82fe4ea5212.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "cd5d289d383ef68f520498dc92e3d671c2f5963a1d9f14c7dbdeb15156357782", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "96393c2797d48908a209ac5ec0599831be6dda63ed8c7f02bb4491be5bee1acb", "generation_seconds_batch": 16.6296143559739, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1703567504882812, "target_z": 1.140625113483912, "ce": 6.268836498260498, "genuineness": 4.494794845581055, "identity": 0.3684733510017395, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/cd5d289d383ef68f520498dc92e3d671c2f5963a1d9f14c7dbdeb15156357782.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "ce0ee2235b15fabc1bf741224036b9011665f3d09db56e012979c9d20dcdbe4b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2f3ae039f3c0a87d1342c85c6a714d4cdef61c84441759684d9f1c16cc587ea5", "generation_seconds_batch": 21.215285453014076, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0807387828826904, "target_z": -0.37499995676803355, "ce": 6.284721851348877, "genuineness": 4.362998008728027, "identity": 0.6225104331970215, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ce0ee2235b15fabc1bf741224036b9011665f3d09db56e012979c9d20dcdbe4b.mp3" }, { "id": "heldout-fr-0", "language": "fr", "sentences": [ "Ton message est arrivé après mon départ.", "Je l’ai relu... puis j’ai rangé mon téléphone.", "Dis-moi quand nous pourrons parler." ], "text": "Ton message est arrivé après mon départ. Je l’ai relu... puis j’ai rangé mon téléphone. Dis-moi quand nous pourrons parler.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-0", "key": "ce26da6ed68a8b0d23444334e35235d29424a289832c10efa6bf3237e06949ee", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Natural clear speech.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5fcbefd004da49d2e79eccb3fa1e01ec04db12b66e0a36e36ac6087a9f5d053b", "generation_seconds_batch": 15.552651091013104, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4719510078430176, "target_z": 1.359375059443954, "ce": 5.34748649597168, "genuineness": 4.356721878051758, "identity": 0.6399667263031006, "hypothesis": "", "wer": 1.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ce26da6ed68a8b0d23444334e35235d29424a289832c10efa6bf3237e06949ee.mp3" }, { "id": "heldout-en-0", "language": "en", "sentences": [ "Your message arrived after I had left.", "I read it twice... before putting the phone away.", "Let me know when we can talk." ], "text": "Your message arrived after I had left. I read it twice... before putting the phone away. Let me know when we can talk.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-0", "key": "ce31d9ba3d7aaf9e4ce075a05b11cab632f60a41e19649ef776cf5cedf894243", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Your message arrived after I had left.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Let me know when we can talk.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Your message arrived after I had left.\n(Natural clear speech.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Natural clear speech.) [3.2 seconds duration] Let me know when we can talk.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5d4a754443974b60db7b7a0972b8ef554b269834329db19e306d8590e82f9e26", "generation_seconds_batch": 16.76012031896971, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.278069019317627, "target_z": 1.2187501188879077, "ce": 5.895530700683594, "genuineness": 4.369274139404297, "identity": 0.3441299796104431, "hypothesis": "Your message arrived after I had left. I read it twice before putting the phone away. Let me know when we can talk.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ce31d9ba3d7aaf9e4ce075a05b11cab632f60a41e19649ef776cf5cedf894243.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "ce42ea7bfbf3383b94a0b2ab03418b3e26a6869bd234ffdebd6c3f2f4327973d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7706ff58baf054a3dcc6c807b95803d676c2fdb3ecf61f75afec6ff6ac016f37", "generation_seconds_batch": 25.028561326907948, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6766042709350586, "target_z": 1.5078125351259728, "ce": 6.117925643920898, "genuineness": 4.595211505889893, "identity": 0.773655116558075, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ce42ea7bfbf3383b94a0b2ab03418b3e26a6869bd234ffdebd6c3f2f4327973d.mp3" }, { "id": "heldout-fr-1", "language": "fr", "sentences": [ "Les clés sont sur la table de la cuisine.", "J’ai regardé dans le tiroir... puis près de la fenêtre.", "Prends-les quand tu partiras." ], "text": "Les clés sont sur la table de la cuisine. J’ai regardé dans le tiroir... puis près de la fenêtre. Prends-les quand tu partiras.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-1", "key": "cea01b3deb479f9bf345325fd30fe5d4de69ad303cba2cdaa6e467be0706803e", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Prends-les quand tu partiras.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Natural clear speech.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Prends-les quand tu partiras.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "e54fdd4be4a73586cb9456d0fa1da91e2f612b00b4a967fa5e3853ea8d40400d", "generation_seconds_batch": 15.476682175998576, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.161309242248535, "target_z": 1.859374886516088, "ce": 5.780362129211426, "genuineness": 4.570107460021973, "identity": 0.7860268354415894, "hypothesis": "Les clés sont sur la table de la cuisine. J'ai regardé dans le tiroir, puis près de la fenêtre. Prends-les quand tu partiras.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/cea01b3deb479f9bf345325fd30fe5d4de69ad303cba2cdaa6e467be0706803e.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "ceb9ef60e03cdf2468e52117f69e39f611e957117957a465c015ed037cfeaaae", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2e1ad7b16698279e4fea03fe5de9646cb377b4602becac3d76679f35fc42fc5a", "generation_seconds_batch": 23.348486786009744, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.805858850479126, "target_z": 1.6015624378540483, "ce": 5.7128496170043945, "genuineness": 4.557555198669434, "identity": 0.5043938755989075, "hypothesis": "Die Tür ist noch offen. Ich hab deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.05263157894736842, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ceb9ef60e03cdf2468e52117f69e39f611e957117957a465c015ed037cfeaaae.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "cf4ed3ca06287a012810131eed39621e337908651a105020ebac773c2938b6e8", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "24435d44613c92456a94034e51670fa44334aa50c30900a86a0f93ffedbd086d", "generation_seconds_batch": 16.690557966940105, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.20094633102417, "target_z": 0.43750006484794973, "ce": 6.197352409362793, "genuineness": 4.507347106933594, "identity": 0.5068812966346741, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/cf4ed3ca06287a012810131eed39621e337908651a105020ebac773c2938b6e8.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "cf5b9817b771506a5436001cb352cbc49ede7d968576a52df27301f043f49d4c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1fb8b6f7cf700b41406b16e6e8f32099c277d49fa61e3b123e3faa401a30417f", "generation_seconds_batch": 23.24093237204943, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.945884943008423, "target_z": 1.7031250486359624, "ce": 5.71682071685791, "genuineness": 4.532451152801514, "identity": 0.7753991484642029, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. We've regarded ensemble what contained.", "wer": 0.3157894736842105, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/cf5b9817b771506a5436001cb352cbc49ede7d968576a52df27301f043f49d4c.mp3" }, { "id": "heldout-de-3", "language": "de", "sentences": [ "Jemand hat den kleinen blauen Stuhl umgestellt.", "Er stand hier... neben dem hohen Bücherregal.", "Wir stellen ihn nach dem Essen zurück." ], "text": "Jemand hat den kleinen blauen Stuhl umgestellt. Er stand hier... neben dem hohen Bücherregal. Wir stellen ihn nach dem Essen zurück.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-3", "key": "cf68e6506caaaad7e7371e72de9d5417fef89211300487ff83acd2484eaab510", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "dc91342af6b3671985295455c66526870f59696e598d4acfa9f1600a1efc9723", "generation_seconds_batch": 15.478994615958072, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8737702369689941, "target_z": 0.20019530473175604, "ce": 6.356205940246582, "genuineness": 4.494794845581055, "identity": 0.576418936252594, "hypothesis": "Jemand hat den kleinen blauen Stuhl umgestellt. Er stand hier neben dem hohen Bücherregal. Wir stellen ihn nach dem Essen zurück.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/cf68e6506caaaad7e7371e72de9d5417fef89211300487ff83acd2484eaab510.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "cf8d3b1e860d24a809e08c722902a65902bf9f22bda150a079a4cc2f82d7d720", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "783d97890274680739bdd498f98907c2f9f400371680b066de0e8a80b2190bc2", "generation_seconds_batch": 23.839920007972978, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.687375545501709, "target_z": 1.5156250702519456, "ce": 6.205295085906982, "genuineness": 4.632867813110352, "identity": 0.7393728494644165, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/cf8d3b1e860d24a809e08c722902a65902bf9f22bda150a079a4cc2f82d7d720.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "cff5c8c840c53a2bbbedf8be7070bada497f044e7ed23546b33148f712ebf794", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "93b295e46f35e26b3b5e48c086577756b5dd5931c32ffc1a1291c53ad23fe793", "generation_seconds_batch": 24.083597186021507, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1488142013549805, "target_z": 1.1250000432319665, "ce": 4.998009204864502, "genuineness": 3.8671910762786865, "identity": 0.3929640054702759, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás lento.", "wer": 0.06666666666666667, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/cff5c8c840c53a2bbbedf8be7070bada497f044e7ed23546b33148f712ebf794.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "d040ef2c910d332864ea05dbd41732f47511d69ce3029f6e2c80fae6095c6f09", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "05d24aa253b9b0d92275c13702ea51423901d96cc99344c55275734e704282e2", "generation_seconds_batch": 20.989678737008944, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1272716522216797, "target_z": 1.1093749729800209, "ce": 6.324435234069824, "genuineness": 4.469690799713135, "identity": 0.27305081486701965, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d040ef2c910d332864ea05dbd41732f47511d69ce3029f6e2c80fae6095c6f09.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "d074ee8a36747d5df5b48ace266a18fa3d0916c8264628a1091ddb16eea7df74", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3c316ddb11d0be11692c2dee3465b24efc1671b1a528c07d33721c0d9845674f", "generation_seconds_batch": 22.698339018970728, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.364238739013672, "target_z": 1.281250054039958, "ce": 6.141753673553467, "genuineness": 3.9927117824554443, "identity": 0.25369974970817566, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d074ee8a36747d5df5b48ace266a18fa3d0916c8264628a1091ddb16eea7df74.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "d07b16cd578eb0afcf0cc6e1be21f72de2ec5d6f63925f99044227401fd8cec0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9de13e1a90f6b3e8558379069c836731d863151e6fb085a520738fbffa107ae6", "generation_seconds_batch": 16.601274695945904, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7950878143310547, "target_z": 1.5937500756559413, "ce": 5.95112943649292, "genuineness": 4.595211505889893, "identity": 0.5204809308052063, "hypothesis": "Encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.058823529411764705, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d07b16cd578eb0afcf0cc6e1be21f72de2ec5d6f63925f99044227401fd8cec0.mp3" }, { "id": "heldout-de-4", "language": "de", "sentences": [ "Ich schrieb die Nummer auf den Umschlag.", "Lies sie genau... und prüfe die letzten beiden Ziffern.", "Ruf an, wenn du Zeit hast." ], "text": "Ich schrieb die Nummer auf den Umschlag. Lies sie genau... und prüfe die letzten beiden Ziffern. Ruf an, wenn du Zeit hast.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-4", "key": "d0a19f545833f0acd63ed3061cff590bcb0496cfe5fb7a000ea9b392ea079277", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Natural clear speech.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Natural clear speech.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b940b453e328ddf90ff84fd743a4a2fe2eb967c0151e493d597a589bfeab09a0", "generation_seconds_batch": 16.94337453308981, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.814906120300293, "target_z": 0.8828124918940062, "ce": 6.078212261199951, "genuineness": 4.419482707977295, "identity": 0.7696083784103394, "hypothesis": "Ich schrieb die Nummer auf den Umschlag, lies sie genau und prüfe die letzten beiden Ziffern. Ruf an, wenn du Zeit hast.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d0a19f545833f0acd63ed3061cff590bcb0496cfe5fb7a000ea9b392ea079277.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "d0ed34783169e35d356fad32b55c337b45b4f02846440bce4b23ab14390c8813", "bucket": "vn_S_ASMR_reg", "phase": "wording", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "fceb4fabec6b53068c51a0462eacc49c721da47450911fad98ad255eb8c53867", "generation_seconds_batch": 17.095417739939876, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.114776372909546, "target_z": 0.37499995676803355, "ce": 6.316492557525635, "genuineness": 4.6705241203308105, "identity": 0.4563267230987549, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d0ed34783169e35d356fad32b55c337b45b4f02846440bce4b23ab14390c8813.mp3" }, { "id": "heldout-en-4", "language": "en", "sentences": [ "I wrote the number on the envelope.", "Read it carefully... and check the last two digits.", "Call me when you have a moment." ], "text": "I wrote the number on the envelope. Read it carefully... and check the last two digits. Call me when you have a moment.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-4", "key": "d11caac31fa3732e2517803823f1b039213fb31e9438619fd83d8b232c894b7d", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I wrote the number on the envelope.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Call me when you have a moment.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I wrote the number on the envelope.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Call me when you have a moment.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "74f9a9ea8a623b4a93b332c48cf08673efda73a0ec095b06c7b466b13cbf3158", "generation_seconds_batch": 17.358769585960545, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3957971334457397, "target_z": -0.14648434190052567, "ce": 6.475345611572266, "genuineness": 4.231201171875, "identity": 0.28445303440093994, "hypothesis": "I wrote the number on the envelope. Read it carefully, and check the last two digits. Call me when you have a moment.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d11caac31fa3732e2517803823f1b039213fb31e9438619fd83d8b232c894b7d.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "d13ed4b9c9e722856f47f11a1cd49da84fde6031799953a2b1d59bf45c6506c1", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "4e129d7040c6d12f5c5ba2debf3acf746d35f5dc60b31c6f8594eb3060af5704", "generation_seconds_batch": 23.24093237204943, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7089178562164307, "target_z": 1.5312499675760252, "ce": 6.09409761428833, "genuineness": 4.532451152801514, "identity": 0.7448815107345581, "hypothesis": "J'ai trouvé la petite boîte, elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d13ed4b9c9e722856f47f11a1cd49da84fde6031799953a2b1d59bf45c6506c1.mp3" }, { "id": "heldout-en-4", "language": "en", "sentences": [ "I wrote the number on the envelope.", "Read it carefully... and check the last two digits.", "Call me when you have a moment." ], "text": "I wrote the number on the envelope. Read it carefully... and check the last two digits. Call me when you have a moment.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-4", "key": "d1754f0e1f75c6ac79fd9ea5b7f9d9e8d7ea3ee34f318fbaacd630f4dfbce61c", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I wrote the number on the envelope.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Call me when you have a moment.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I wrote the number on the envelope.\n(Natural clear speech.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Natural clear speech.) [3.2 seconds duration] Call me when you have a moment.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "510a599828614b4a41fabc2821ddcf3de9ade9e4d08888238b04ec26b9861cb6", "generation_seconds_batch": 17.297314282972366, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.847219944000244, "target_z": 0.9062500972719246, "ce": 6.181467056274414, "genuineness": 4.15588903427124, "identity": 0.38587918877601624, "hypothesis": "I wrote the number on the envelope. Read it carefully and check the last two digits. Call me when you have a moment.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d1754f0e1f75c6ac79fd9ea5b7f9d9e8d7ea3ee34f318fbaacd630f4dfbce61c.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "d1821f72857081f087737413d1d06b2a3bbfd887c99e4c0782f7a3eb013eae6e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 130, "seconds_generated": 10.4, "mp3_sha256": "e7e976fbb41cf78b6de5c70ae56138f65daf00cdc769c077f1e6cdb2ce8a8608", "generation_seconds_batch": 23.259458995074965, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7950878143310547, "target_z": 1.5937500756559413, "ce": 5.903473377227783, "genuineness": 4.419482707977295, "identity": 0.3286708891391754, "hypothesis": "I found the little box. It was beside the old window. We can look inside.", "wer": 0.0625, "duration_error_s": 0.20000000000000107, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d1821f72857081f087737413d1d06b2a3bbfd887c99e4c0782f7a3eb013eae6e.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "d1912caf70613d179632c9b172f40bda5a6eb64ec8a4cb90678e7e4162171f81", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "83afcc2c96e9cfc7f9c9bd985daf7f65442e1c8a3f0a2f5e872c1014eb7ae872", "generation_seconds_batch": 24.058624504017644, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7071938514709473, "target_z": 0.8046874864900104, "ce": 6.1020402908325195, "genuineness": 4.507347106933594, "identity": 0.7292249798774719, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d1912caf70613d179632c9b172f40bda5a6eb64ec8a4cb90678e7e4162171f81.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "d19b4986411a6284084d2dab1eded787a103efa8be668d813255ecd305cc8639", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8e5d1f187b338f579a7bdb0c19b1f1941a2d82bcbbcb9e15a94a09a03c1c1fa6", "generation_seconds_batch": 23.83307847904507, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3319249153137207, "target_z": 1.25781244866204, "ce": 6.205295085906982, "genuineness": 4.218649387359619, "identity": 0.3283393979072571, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d19b4986411a6284084d2dab1eded787a103efa8be668d813255ecd305cc8639.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "d1be474e5bebecc4329468423ef77b87183ec7b32b94504c6593d72186c742cd", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7fbae77a819c44824dff38a2d6a75efc547a2fccf8bf5561097a1c0e3c9fd481", "generation_seconds_batch": 16.971309890039265, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.741231679916382, "target_z": 1.5546875729539433, "ce": 5.982900142669678, "genuineness": 4.6956281661987305, "identity": 0.6097972393035889, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d1be474e5bebecc4329468423ef77b87183ec7b32b94504c6593d72186c742cd.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "d1c68affcc2b6afd407608a6de20c369ef4e5d75a93e99ae5fb298958262a633", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9db1071fff56d964dc11d8ff898e5029e6c1d2aa7ea76a87185cdbfc3c29b18a", "generation_seconds_batch": 24.162064735079184, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.06436824798584, "target_z": 1.789062416238065, "ce": 5.788304805755615, "genuineness": 4.6203155517578125, "identity": 0.4951668679714203, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay.", "wer": 0.058823529411764705, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d1c68affcc2b6afd407608a6de20c369ef4e5d75a93e99ae5fb298958262a633.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "d1d3091aa6edb856dd805de5ebd35362c053673e5a0b39a1165794f1c12092bf", "bucket": "vn_S_ASMR_reg", "phase": "wording", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c492f0c09b835053a651909ff6ed27922f073797790e3ec4503b3cfe9758f3e2", "generation_seconds_batch": 17.16178419603966, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9837976694107056, "target_z": -0.4453125135099895, "ce": 6.078212261199951, "genuineness": 4.337893962860107, "identity": 0.7106397151947021, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d1d3091aa6edb856dd805de5ebd35362c053673e5a0b39a1165794f1c12092bf.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "d1dda466da04291474564bb7dce5464e8d5457e6cee837d61429960a4060ce94", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "61e6828cadf44881ec2a83ea6b50554ee747938234241ed33dcfcae9958e0344", "generation_seconds_batch": 22.968272380996495, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.637180805206299, "target_z": 0.7539061810990534, "ce": 6.133810997009277, "genuineness": 4.356721878051758, "identity": 0.6904963850975037, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d1dda466da04291474564bb7dce5464e8d5457e6cee837d61429960a4060ce94.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "d1e9394c092881e2537967f14bcc06904ab8f0260204c55a54ff6957c5558187", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "417a3fb9d398ec2719217a6dc2829fb2623c83822640223fdde01fa5e3eca93e", "generation_seconds_batch": 24.249534575035796, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.742955684661865, "target_z": 2.2812500540399583, "ce": 5.855817794799805, "genuineness": 4.607763767242432, "identity": 0.7573095560073853, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d1e9394c092881e2537967f14bcc06904ab8f0260204c55a54ff6957c5558187.mp3" }, { "id": "heldout-es-1", "language": "es", "sentences": [ "Las llaves están en la mesa de la cocina.", "Miré en el cajón... y luego junto a la ventana.", "Llévatelas cuando te vayas." ], "text": "Las llaves están en la mesa de la cocina. Miré en el cajón... y luego junto a la ventana. Llévatelas cuando te vayas.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-1", "key": "d2319f0ba4d4a13f099c61fd2524d2cd47bb09bca092815ed8a29e7f5b0c527c", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Llévatelas cuando te vayas.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Llévatelas cuando te vayas.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e134f8e0f365b5b93422b6f8e1da05dbeeebf800b62ec506bd31824b5a592f9f", "generation_seconds_batch": 17.944635855033994, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8306853771209717, "target_z": 0.16894533715573087, "ce": 6.157639026641846, "genuineness": 4.231201171875, "identity": 0.6866028308868408, "hypothesis": "Las llaves están en la mesa de la cocina. Miré en el cajón y luego junto a la ventana. Llévatelas cuando te vayas.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d2319f0ba4d4a13f099c61fd2524d2cd47bb09bca092815ed8a29e7f5b0c527c.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "d231af8e5ef12aa5c927922f8ccd22b10af07ca113e0ab7a8b8e229340804876", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3acec1d162171897fb7acf09682165b3197b44787fdcbddf155f7561a2da1d67", "generation_seconds_batch": 22.24161526595708, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7089178562164307, "target_z": 1.5312499675760252, "ce": 6.268836498260498, "genuineness": 4.268857479095459, "identity": 0.31843364238739014, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d231af8e5ef12aa5c927922f8ccd22b10af07ca113e0ab7a8b8e229340804876.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "d278a2195d9cfd1ebfcd5d52d3be2789822a1ae9d4a846ed2c70512296c6134b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1c551cf52c4639735bb0f20729b681a61cf9e30cf608b68b731616b45c3825d1", "generation_seconds_batch": 19.295539007987827, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7089178562164307, "target_z": 1.5312499675760252, "ce": 5.426913261413574, "genuineness": 4.444586753845215, "identity": 0.38216322660446167, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d278a2195d9cfd1ebfcd5d52d3be2789822a1ae9d4a846ed2c70512296c6134b.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "d283cc1793873b2e7e03409765ad0a5c928077d0c91955c74481897b83641ac1", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6f748152621c87a71735d2e1a0ce3ac4ebd2d2afc96772ac2bc8a88cb30aaba7", "generation_seconds_batch": 25.116734603070654, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3857812881469727, "target_z": 1.2968751242919037, "ce": 6.141753673553467, "genuineness": 4.65797233581543, "identity": 0.8377482891082764, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d283cc1793873b2e7e03409765ad0a5c928077d0c91955c74481897b83641ac1.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "d2ca46efe0cd28760e8243c6f56e20d454e5df1a1a2c43be237de092ef5c7859", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "0d0dd6788cd76cc5944ed9dfaa150b2c0c8321e9f6844f6efa7389619a5cbc4e", "generation_seconds_batch": 24.02223238791339, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.945884943008423, "target_z": 1.7031250486359624, "ce": 5.260117053985596, "genuineness": 4.457139015197754, "identity": 0.4355190396308899, "hypothesis": "He encontrado la cajita, estaba junto a la ventana vieja, podemos mirar juntos lo que hay.", "wer": 0.058823529411764705, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d2ca46efe0cd28760e8243c6f56e20d454e5df1a1a2c43be237de092ef5c7859.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "d33687e9d2aa02fefadfec2c49b8465a6491109876c00e2d79401b7d0d9b5d6c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "599bebca42c62fee8b15810a237648c462e307c08425e13fc846dd4b018c8a5d", "generation_seconds_batch": 24.94929073692765, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7089178562164307, "target_z": 1.5312499675760252, "ce": 5.200547218322754, "genuineness": 4.381826400756836, "identity": 0.4427720010280609, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d33687e9d2aa02fefadfec2c49b8465a6491109876c00e2d79401b7d0d9b5d6c.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "d3c59aa9755c25ea71be67505a920c00ee9495a80c3b7ac68872e043933b003a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "bd6a93f233dfcd8dd4bb966a7c055d8066cd29b1a0c021e3bc726b869640e752", "generation_seconds_batch": 22.549199555069208, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2134413719177246, "target_z": 1.1718749081320712, "ce": 5.8875885009765625, "genuineness": 4.268857479095459, "identity": 0.43747153878211975, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás.", "wer": 0.06666666666666667, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d3c59aa9755c25ea71be67505a920c00ee9495a80c3b7ac68872e043933b003a.mp3" }, { "id": "heldout-fr-4", "language": "fr", "sentences": [ "J’ai écrit le numéro sur l’enveloppe.", "Lis-le bien... et vérifie les deux derniers chiffres.", "Appelle-moi quand tu auras un moment." ], "text": "J’ai écrit le numéro sur l’enveloppe. Lis-le bien... et vérifie les deux derniers chiffres. Appelle-moi quand tu auras un moment.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-4", "key": "d3fab0ead1fc9dcf899f59bddb0c9d4b54373ba1829b6dc49f77fc4027ce1740", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "430c58d9b9a907856442239c7a3b250ab880671c66f13236910c46331901c9f3", "generation_seconds_batch": 17.850444269017316, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.605836033821106, "target_z": 0.005859401344479572, "ce": 5.990842819213867, "genuineness": 4.293961524963379, "identity": 0.7483785152435303, "hypothesis": "J'ai écrit le numéro sur l'enveloppe. Lis-le bien et vérifie les deux derniers chiffres. Appelle-moi quand tu auras un moment.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d3fab0ead1fc9dcf899f59bddb0c9d4b54373ba1829b6dc49f77fc4027ce1740.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "d48556e8198fb37c21b7f7a223fe91c2dc4e547a43de9860e4b6582156916ab0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c96e51c0f078b16711d16c8bb7036fab26bb694e391cb4f6c1e5e802366fa490", "generation_seconds_batch": 23.116177136078477, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.161309242248535, "target_z": 1.859374886516088, "ce": 6.054384231567383, "genuineness": 4.7458367347717285, "identity": 0.5018056631088257, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d48556e8198fb37c21b7f7a223fe91c2dc4e547a43de9860e4b6582156916ab0.mp3" }, { "id": "heldout-de-1", "language": "de", "sentences": [ "Die Schlüssel liegen auf dem Küchentisch.", "Ich suchte in der Schublade... dann neben dem Fenster.", "Nimm sie mit, wenn du gehst." ], "text": "Die Schlüssel liegen auf dem Küchentisch. Ich suchte in der Schublade... dann neben dem Fenster. Nimm sie mit, wenn du gehst.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-1", "key": "d487936a9f500603a28de37ccb2314b2b7c7adf8c978a04379e6e485ab4c7b3b", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Natural clear speech.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Natural clear speech.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "bd9f6a284d2bada3e04d939816816bf10442e114d7ec27b28c73cfb6234e1576", "generation_seconds_batch": 16.35157260799315, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9028000831604004, "target_z": 1.6718750810599372, "ce": 6.022613525390625, "genuineness": 4.582659721374512, "identity": 0.525098443031311, "hypothesis": "Die Schlüssel liegen auf dem Küchentisch. Ich suchte in der Schublade, daneben dem Fenster. Nimm sie mit, wenn du gehst.", "wer": 0.09523809523809523, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d487936a9f500603a28de37ccb2314b2b7c7adf8c978a04379e6e485ab4c7b3b.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "d57c5b839ff043d19729b527b834b49f58b1e104687e3d8b0d0e9c8684150dc6", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "b0ea7ff65030526779efdb969afa87f8947d50a6517921ee8dbe857b6d40841b", "generation_seconds_batch": 23.119388901046477, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8041348457336426, "target_z": 0.8749999567680335, "ce": 6.205295085906982, "genuineness": 4.68307638168335, "identity": 0.6977099180221558, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d57c5b839ff043d19729b527b834b49f58b1e104687e3d8b0d0e9c8684150dc6.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "d59d231087d30c703239ecb90e01681ce754aedb73b8be56c8b5fecfa224e518", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "111f5fc574af573ebdb1e576d46c3f34098976159244b78540facbf580fb8ca5", "generation_seconds_batch": 23.008963654050604, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.053597450256348, "target_z": 1.781250226967824, "ce": 5.629451274871826, "genuineness": 4.406930446624756, "identity": 0.5569595098495483, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d59d231087d30c703239ecb90e01681ce754aedb73b8be56c8b5fecfa224e518.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "d5fc48e3360910a75af8c24c9da9cd5f7d2e560a331f93533a89820c2ae5d5d5", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4e88398dc1c2b3e56701c480fbb858100b016f74bbd27be58edafd0ba589adbd", "generation_seconds_batch": 24.725788276991807, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7089178562164307, "target_z": 1.5312499675760252, "ce": 6.181467056274414, "genuineness": 4.557555198669434, "identity": 0.38698118925094604, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d5fc48e3360910a75af8c24c9da9cd5f7d2e560a331f93533a89820c2ae5d5d5.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "d60768f9df33af48dff2e7bb7dae9ebaa38473c25888a9f476c853f74281be6a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1c2c919c917719c5005d31c8883c09f6516e2fc89f56cc4710cd06c17d7488f6", "generation_seconds_batch": 25.08338487497531, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3211536407470703, "target_z": 1.249999913536067, "ce": 6.205295085906982, "genuineness": 4.6203155517578125, "identity": 0.6593821048736572, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d60768f9df33af48dff2e7bb7dae9ebaa38473c25888a9f476c853f74281be6a.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "d640d6e71c548aedf824f7104e29ba52d8d5cedf29eecacd13e2905c705eafd9", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e58e9a22ea3c8a0d2bd2f6c0b167e5715706a4967d89d9d7c517321e5a1041ce", "generation_seconds_batch": 24.410459173028357, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1874823570251465, "target_z": 0.4277344824044167, "ce": 6.062326908111572, "genuineness": 4.532451152801514, "identity": 0.6064243316650391, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d640d6e71c548aedf824f7104e29ba52d8d5cedf29eecacd13e2905c705eafd9.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "d660d2f4fd5ffe16f24dc02708af3f8f1e8974316d0f1bda78a37a39a12f32ba", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.0, "cfg": 2.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "da09b1f393035b63413b58bd7e611db24c7b58a0bfea7a516865e3fe014ed8e6", "generation_seconds_batch": 16.5183989610523, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.773545265197754, "target_z": 1.578125005403996, "ce": 6.157639026641846, "genuineness": 4.758388519287109, "identity": 0.7535256743431091, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d660d2f4fd5ffe16f24dc02708af3f8f1e8974316d0f1bda78a37a39a12f32ba.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "d687254546978165541fc551d87aea04bc1fc506c7e41f7f09733adc072e7a90", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ef2077152cf290bb905948e62b40839987ca38d156d0ea49cd0e2c5202faf46b", "generation_seconds_batch": 25.65067368594464, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5402398109436035, "target_z": 0.6835937108210304, "ce": 6.483288764953613, "genuineness": 4.457139015197754, "identity": 0.3149817883968353, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d687254546978165541fc551d87aea04bc1fc506c7e41f7f09733adc072e7a90.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "d6dd70463833ea79e83b50bb7fedc41bc581f0beddac6afdcfa4f6cf91868d2e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "948350900613016e18cc375708237d44c97002e885d19a8d1498bb6988376713", "generation_seconds_batch": 17.675301842973568, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0097570419311523, "target_z": 0.29882817160946384, "ce": 6.4515180587768555, "genuineness": 4.68307638168335, "identity": 0.3163634240627289, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d6dd70463833ea79e83b50bb7fedc41bc581f0beddac6afdcfa4f6cf91868d2e.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "d889f45de0724030404adc1ab51961b2ebb7e8d6b8064d84a04f3ac494c65889", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "91975d3c719fef92a51a6cb27070f3f79b7f0157dfa2c2281dda2d0c2c62c105", "generation_seconds_batch": 15.723736896063201, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8812575340270996, "target_z": 1.6562500108079916, "ce": 5.295859336853027, "genuineness": 4.381826400756836, "identity": 0.475044310092926, "hypothesis": "Encontrado la cajita, estaba junto a la ventana vieja, podemos mirar juntos lo que hay dentro.", "wer": 0.058823529411764705, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d889f45de0724030404adc1ab51961b2ebb7e8d6b8064d84a04f3ac494c65889.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "d989bb4321a05a045fe3b596e64c701b95616648b7f50ffbe3ba316fc122bc3d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3d32352699a2179203a919cc6f2ed1f2fd0ef18c2a3743560a06e4f44aec225d", "generation_seconds_batch": 23.898921641986817, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7377835512161255, "target_z": 0.10156252431798114, "ce": 6.348263263702393, "genuineness": 4.055472373962402, "identity": 0.5072746872901917, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d989bb4321a05a045fe3b596e64c701b95616648b7f50ffbe3ba316fc122bc3d.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "d99134e92f882013fff639b2c5ae800c9c0f3de019b062610ff5c60e9e863916", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "54f4cf9d624c697ec4e5132e1e7808b87f1ad90170b390fcef240a9aefe3668b", "generation_seconds_batch": 21.8217074990971, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5581207275390625, "target_z": 1.421874994596004, "ce": 5.816103935241699, "genuineness": 4.632867813110352, "identity": 0.5675632357597351, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/d99134e92f882013fff639b2c5ae800c9c0f3de019b062610ff5c60e9e863916.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "db0df57c6682c5423ebffe46975d330da09cba2128785bcb168c53ae1dc7d2b9", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7a91930b90c0c22c576d694fb5d1198e0c5e04be7cb5b48fa0a28048e69f51fc", "generation_seconds_batch": 24.118244329933077, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7323979139328003, "target_z": 0.09765625675499476, "ce": 6.086154937744141, "genuineness": 4.406930446624756, "identity": 0.5446997284889221, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/db0df57c6682c5423ebffe46975d330da09cba2128785bcb168c53ae1dc7d2b9.mp3" }, { "id": "heldout-en-0", "language": "en", "sentences": [ "Your message arrived after I had left.", "I read it twice... before putting the phone away.", "Let me know when we can talk." ], "text": "Your message arrived after I had left. I read it twice... before putting the phone away. Let me know when we can talk.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-0", "key": "db192910e9ae67de251fb2a990940eab881f66f627c78b50b638ce61f5a63679", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Your message arrived after I had left.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Let me know when we can talk.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Your message arrived after I had left.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Let me know when we can talk.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0d2620d4e2a763873fd75fbd135623075d907fdc71244d151ea1fb597f005197", "generation_seconds_batch": 16.32647258997895, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2046078443527222, "target_z": -0.2851562351390115, "ce": 6.427689552307129, "genuineness": 4.306513786315918, "identity": 0.4576208293437958, "hypothesis": "Your message arrived after I had left. I read it twice before putting the phone away. Let me know when we can talk.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/db192910e9ae67de251fb2a990940eab881f66f627c78b50b638ce61f5a63679.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "db74795bed3408440cefef54122fce4a3dde7e6407ee08eb49d95639f9e6aea9", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "dd751d37b4057ebab13582aa9d461e1f1b0ac2c8c0bc609968f10c41db5e77ee", "generation_seconds_batch": 24.63505137595348, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9889698028564453, "target_z": 1.7343750162119875, "ce": 6.038498878479004, "genuineness": 4.6203155517578125, "identity": 0.738540530204773, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/db74795bed3408440cefef54122fce4a3dde7e6407ee08eb49d95639f9e6aea9.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "dbd49a9aae36fe460e9cab76afda0a1d26bf6afc8f2e1e569f4cafb87fe9088c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "1f5b89645fb04f87f76a392c1a9d0ce80ce08467db09d132b9feef0980596cb6", "generation_seconds_batch": 20.402951620984823, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.547349452972412, "target_z": 1.4140624594700315, "ce": 5.816103935241699, "genuineness": 4.331617832183838, "identity": 0.6459414958953857, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hinein.", "wer": 0.0625, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/dbd49a9aae36fe460e9cab76afda0a1d26bf6afc8f2e1e569f4cafb87fe9088c.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "dc8d11c57025612da01607151fb57d5a395b1746536010a3cd3d87073accf39f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "14b3fdb3b21bc86359609f5b7d9b79b8cd54851b22aefd35123d09b15639d743", "generation_seconds_batch": 23.60011624090839, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.182851791381836, "target_z": 1.8749999567680335, "ce": 5.871703147888184, "genuineness": 4.406930446624756, "identity": 0.7082014083862305, "hypothesis": "La porte est encore ouverte, j'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/dc8d11c57025612da01607151fb57d5a395b1746536010a3cd3d87073accf39f.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "dca1ca2171bd2d95b3e403cb8a54367a78df914c5e20252792bba159a1e89f3f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ffbeb7ce1cf317bc0fb015c70a72f774aafc5b70d55cd93f8d04f05a51008b91", "generation_seconds_batch": 15.6414084819844, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.224212646484375, "target_z": 1.179687443258044, "ce": 6.229123115539551, "genuineness": 4.2374773025512695, "identity": 0.47497206926345825, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/dca1ca2171bd2d95b3e403cb8a54367a78df914c5e20252792bba159a1e89f3f.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "dcd2605cc093003d2882e78c11e0af02756b0f3d10f594d306f76f1fc638da5b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2cc2f1899d0480f6eee6bfe5815cfa5d2835164ccfb9d38f7c1c980e19c85d7e", "generation_seconds_batch": 21.955919508007355, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8276149034500122, "target_z": -0.5585937540529968, "ce": 6.09409761428833, "genuineness": 4.425758361816406, "identity": 0.7503504753112793, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd, zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/dcd2605cc093003d2882e78c11e0af02756b0f3d10f594d306f76f1fc638da5b.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "de07f09f3dc04f864206a10a5da75ede2135917e04acab584f59256b3ac6371d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "eee6733e22689143471fd8fd1525dbb94fc6d8ea34c18ac14307ba0274114912", "generation_seconds_batch": 17.228487334912643, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.224212646484375, "target_z": 1.179687443258044, "ce": 6.23706579208374, "genuineness": 4.250029563903809, "identity": 0.639369010925293, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschaffen.", "wer": 0.0625, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/de07f09f3dc04f864206a10a5da75ede2135917e04acab584f59256b3ac6371d.mp3" }, { "id": "heldout-en-0", "language": "en", "sentences": [ "Your message arrived after I had left.", "I read it twice... before putting the phone away.", "Let me know when we can talk." ], "text": "Your message arrived after I had left. I read it twice... before putting the phone away. Let me know when we can talk.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-0", "key": "de824e876ad93457e292f7990b389a2088b9e3202ed204a3f9fe732623bc8cbf", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Your message arrived after I had left.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Let me know when we can talk.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Your message arrived after I had left.\n(Natural clear speech.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Natural clear speech.) [3.2 seconds duration] Let me know when we can talk.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "fe3700ca8e055184634013049a8ed0bae8db1aa1d43d1a63103411bf7ea40af1", "generation_seconds_batch": 16.76012031896971, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.51503586769104, "target_z": 1.3906250270199791, "ce": 6.197352409362793, "genuineness": 4.432034492492676, "identity": 0.46592411398887634, "hypothesis": "Your message arrived after I had left. I read it twice before putting the phone away. Let me know when we can talk.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/de824e876ad93457e292f7990b389a2088b9e3202ed204a3f9fe732623bc8cbf.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "ded8e106c80ea9cda136734b27a2605d39bd9898cd2b48f5e13594dffa197686", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "81b6e91e7713ef36956d344b4c9ceea180c283a271109479ca80a53a541f0b9b", "generation_seconds_batch": 23.1860186860431, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.153444528579712, "target_z": -0.32226560405951626, "ce": 6.252951145172119, "genuineness": 4.494794845581055, "identity": 0.580099880695343, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ded8e106c80ea9cda136734b27a2605d39bd9898cd2b48f5e13594dffa197686.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "df4a694fb65b015d8285f936f0ae04e6d7598e29f2ce61a16da50e3c38d5a3c3", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2df0ab36af48aa607cc0911948c90aad8ee8fbc9f9e77a927bdc33768cc2fd87", "generation_seconds_batch": 22.549199555069208, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8489439487457275, "target_z": 1.6328125783579392, "ce": 5.6413655281066895, "genuineness": 4.250029563903809, "identity": 0.3755830228328705, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás lista.", "wer": 0.06666666666666667, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/df4a694fb65b015d8285f936f0ae04e6d7598e29f2ce61a16da50e3c38d5a3c3.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "df8d45c2ab9d3765a13c98d72a6eb8e7526bf6029f469fa537f70e72345e6f37", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "470741fcf2a7f565a0ffb74160dfbddfb01d6a307b74b4149ed07184718383df", "generation_seconds_batch": 17.327590133994818, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.95493221282959, "target_z": 0.9843751026759204, "ce": 6.340320587158203, "genuineness": 4.582659721374512, "identity": 0.6008322238922119, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/df8d45c2ab9d3765a13c98d72a6eb8e7526bf6029f469fa537f70e72345e6f37.mp3" }, { "id": "heldout-nl-4", "language": "nl", "sentences": [ "Ik schreef het nummer op de envelop.", "Lees het goed... en controleer de laatste twee cijfers.", "Bel me wanneer je even tijd hebt." ], "text": "Ik schreef het nummer op de envelop. Lees het goed... en controleer de laatste twee cijfers. Bel me wanneer je even tijd hebt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-4", "key": "df91727faaac3e0b01b76cac412ea98971d09fbc67ae29f6744c40fdf572a878", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Natural clear speech.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Natural clear speech.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7d01d2abc7f21bfa9bceb035768fb32e1bf4878bc6915a451afa916598f55bdb", "generation_seconds_batch": 14.909288667025976, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0303306579589844, "target_z": 1.039062502701998, "ce": 5.919358730316162, "genuineness": 4.243753433227539, "identity": 0.6703301668167114, "hypothesis": "Ik schreef het nummer op de envelop. Lees het goed en controleer de laatste twee cijfers. Bel me wanneer je even tijd hebt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/df91727faaac3e0b01b76cac412ea98971d09fbc67ae29f6744c40fdf572a878.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "e02ff8ed2a3368e747308f41aea076c7a2061f7dae0a691ad019de1d8ffcd89e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7e3c504b8ccc79494bd8d171b8fee9a06bc2279977d22d7a4b0c26c9d6c3b5ff", "generation_seconds_batch": 25.12463709106669, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9882144927978516, "target_z": 0.2832031013575183, "ce": 6.348263263702393, "genuineness": 4.595211505889893, "identity": 0.631308913230896, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e02ff8ed2a3368e747308f41aea076c7a2061f7dae0a691ad019de1d8ffcd89e.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "e04b39d3f661ca452dae3484ed9c13b10aa9275893cb10b53d057338ffb7ebe2", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "cdc99653416705c0c39cf89cbb76218cce2002c0c51e11abcc5011e35d435dc9", "generation_seconds_batch": 22.534372791997157, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4719510078430176, "target_z": 1.359375059443954, "ce": 5.9431867599487305, "genuineness": 4.519899368286133, "identity": 0.7077152729034424, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e04b39d3f661ca452dae3484ed9c13b10aa9275893cb10b53d057338ffb7ebe2.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "e14c982d1963b324d428a11bca8b4c0a4a739c402517ada2d744c5da3f60781f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "932f5983d51ce5955211e03eedb9dcce995eae5efe1b1e75432d463059b9d229", "generation_seconds_batch": 23.535912253893912, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.642566680908203, "target_z": 0.7578126215899057, "ce": 5.9431867599487305, "genuineness": 4.118232727050781, "identity": 0.34015166759490967, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e14c982d1963b324d428a11bca8b4c0a4a739c402517ada2d744c5da3f60781f.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "e1e991ead111ffe4e3718b8c7dc67d30dc0b7073a94a58cd143ffe7823f36c00", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "dd6891a53776c7f48431c6d4e6f1fcead751b15b7d7e72afc1635f4e1064f5b0", "generation_seconds_batch": 22.12326556397602, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.171325445175171, "target_z": 0.41601567971545755, "ce": 6.213237762451172, "genuineness": 4.432034492492676, "identity": 0.5707295536994934, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e1e991ead111ffe4e3718b8c7dc67d30dc0b7073a94a58cd143ffe7823f36c00.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "e1eddaa185f89433b439e3146c4da35f6a663d2298e92227093c1033c337dbec", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c2a778508655e6bc47903cf209e6d02710f2e270090b2dc209b129b155f00355", "generation_seconds_batch": 23.523478726041503, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.576215147972107, "target_z": -0.01562498378801257, "ce": 5.959072113037109, "genuineness": 4.149612903594971, "identity": 0.5673669576644897, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e1eddaa185f89433b439e3146c4da35f6a663d2298e92227093c1033c337dbec.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "e1f14cfbe7505ff065f462383b8847eff0f4fe1cb878e29285588f2c84eebaf3", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b5a281b01b63376a1cfca1159b8fab6ea67018d0ac51b1a9a7957902b0033f8d", "generation_seconds_batch": 23.912861598073505, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.562751054763794, "target_z": -0.025390652695478525, "ce": 6.086154937744141, "genuineness": 4.369274139404297, "identity": 0.5940815806388855, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e1f14cfbe7505ff065f462383b8847eff0f4fe1cb878e29285588f2c84eebaf3.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "e20393f170c7670c22300634099736b5f36c16f3ec1ace19c9f678bbdaff8804", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1cf35fdbd62a5f6e786d24388c9c87f5af25a61605ad12b4685664bf3d419375", "generation_seconds_batch": 17.535287783946842, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5581207275390625, "target_z": 1.421874994596004, "ce": 5.895530700683594, "genuineness": 4.557555198669434, "identity": 0.6961311101913452, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e20393f170c7670c22300634099736b5f36c16f3ec1ace19c9f678bbdaff8804.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "e242906610172657347f2c72634a7b34ec0db50d92d4510fdc257a03f9a58d10", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "39a12e40c0043e49f9ac46199ca2dc29edc71c5f5dfbeb01b0c37e3c2a2b3dc0", "generation_seconds_batch": 26.507710720063187, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7916396856307983, "target_z": 0.14062502701997906, "ce": 6.300607204437256, "genuineness": 3.911123275756836, "identity": 0.24772049486637115, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e242906610172657347f2c72634a7b34ec0db50d92d4510fdc257a03f9a58d10.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "e27aec696717b19dabe492b18149e8f49def706d46929cb5f8a287d45388569e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4942334cde5362dd4dbbac1cc0b2aad5aa9965b5b6a74e0330c1846f032ee65b", "generation_seconds_batch": 22.9791196139995, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5258069038391113, "target_z": 1.398437389218086, "ce": 5.5539960861206055, "genuineness": 4.137060642242432, "identity": 0.35964059829711914, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo.", "wer": 0.13333333333333333, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e27aec696717b19dabe492b18149e8f49def706d46929cb5f8a287d45388569e.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "e29fc347c42f12fa21a75cdfeeb727876609df192d6993b70c780ca748da1547", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 131, "seconds_generated": 10.48, "mp3_sha256": "5dc7b1cef25d04c3c6de67d4e34ec73c48fdf612befffd78c4e110a9237b7e66", "generation_seconds_batch": 23.870138016995043, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.838172674179077, "target_z": 1.6250000432319665, "ce": 5.11317777633667, "genuineness": 4.074300289154053, "identity": 0.4990198016166687, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás.", "wer": 0.06666666666666667, "duration_error_s": 0.28000000000000114, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e29fc347c42f12fa21a75cdfeeb727876609df192d6993b70c780ca748da1547.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "e2fb16b82e7340386222a18a006edc778fb175a1248871ab2fdc6c4764b1082d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0bababb62c10b271fa23a5cda48c9f79d9c2d4f6f6a6c8a25efb1e9c08d8f92d", "generation_seconds_batch": 22.516669878037646, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7089178562164307, "target_z": 1.5312499675760252, "ce": 6.038498878479004, "genuineness": 4.65797233581543, "identity": 0.783173680305481, "hypothesis": "La porte est encore ouverte j'ai posé ton manteau sur la chaise. Dis moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e2fb16b82e7340386222a18a006edc778fb175a1248871ab2fdc6c4764b1082d.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "e3054bd1bf0948cb449f6556f74274d8aac2904effe3e3b81bfb744647f63200", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "28f71f5e09b87147582ebde19125cb130f1e0d5e45802cf0bdc20851441bc139", "generation_seconds_batch": 24.333596550975926, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.07513952255249, "target_z": 1.7968749513640376, "ce": 6.09409761428833, "genuineness": 4.720732688903809, "identity": 0.821886420249939, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e3054bd1bf0948cb449f6556f74274d8aac2904effe3e3b81bfb744647f63200.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "e367a68aae2bae9ecf90a54f13ecb2e9cb7cb67693a73bc715dc3c2e8c9916e0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "707e183fd0efdaea713dde336a4feb5d66522b17e8bf7e8debeb295a3b348107", "generation_seconds_batch": 25.48447667597793, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.577939033508301, "target_z": 0.7109374108340691, "ce": 6.23706579208374, "genuineness": 4.519899368286133, "identity": 0.4841844141483307, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e367a68aae2bae9ecf90a54f13ecb2e9cb7cb67693a73bc715dc3c2e8c9916e0.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "e37f59927e3a6826a1900ebbbbcd87d2641d2dcd2a18240270ea2f1bb7478181", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0d078cd15e6ae4fa13de333fccd4b40e9a6a336dc0f683657e1a76d091f4be13", "generation_seconds_batch": 24.92359939101152, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7054698467254639, "target_z": 0.0781250054039958, "ce": 6.403861999511719, "genuineness": 4.482243061065674, "identity": 0.7437066435813904, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e37f59927e3a6826a1900ebbbbcd87d2641d2dcd2a18240270ea2f1bb7478181.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "e41d82f571dd7f7e591b202c8882d7e2c4ff537af8331c6f42b7199b301cb8fa", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "1c44bbe0142d8960f102ff5c869dae23998fe9fd284688ee43de585b30f5895e", "generation_seconds_batch": 15.354713826905936, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6442904472351074, "target_z": 1.4843749297480544, "ce": 6.221180438995361, "genuineness": 4.306513786315918, "identity": 0.4146934151649475, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e41d82f571dd7f7e591b202c8882d7e2c4ff537af8331c6f42b7199b301cb8fa.mp3" }, { "id": "heldout-en-4", "language": "en", "sentences": [ "I wrote the number on the envelope.", "Read it carefully... and check the last two digits.", "Call me when you have a moment." ], "text": "I wrote the number on the envelope. Read it carefully... and check the last two digits. Call me when you have a moment.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-4", "key": "e5dda9d46208d6c63888f588d9684d3cb458bab50aeba421b311caf199edec78", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I wrote the number on the envelope.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Call me when you have a moment.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I wrote the number on the envelope.\n(Natural clear speech.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Natural clear speech.) [3.2 seconds duration] Call me when you have a moment.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 130, "seconds_generated": 10.4, "mp3_sha256": "e65aa06050eb8747850552bdab147fa513315db30fa2e5eafcab4402e373b418", "generation_seconds_batch": 16.308117015985772, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3679003715515137, "target_z": 0.5585938405169298, "ce": 5.637393951416016, "genuineness": 3.7416701316833496, "identity": 0.30973100662231445, "hypothesis": "I wrote the number on the envelope. Read it carefully and check the last two digits. Call me when you have a m-", "wer": 0.043478260869565216, "duration_error_s": 0.20000000000000107, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e5dda9d46208d6c63888f588d9684d3cb458bab50aeba421b311caf199edec78.mp3" }, { "id": "heldout-es-3", "language": "es", "sentences": [ "Alguien ha movido la pequeña silla azul.", "Estaba aquí... junto a la estantería alta.", "La colocaremos después de comer." ], "text": "Alguien ha movido la pequeña silla azul. Estaba aquí... junto a la estantería alta. La colocaremos después de comer.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-3", "key": "e684089122770917eb486f01ea45421839eafa6eeacf493f6f0682504ad528b8", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La colocaremos después de comer.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Natural clear speech.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Natural clear speech.) [3.2 seconds duration] La colocaremos después de comer.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8657e0410d9c1649e088dabf4e4ff541411f8c166bfbf1759a762b2e5bc1fa67", "generation_seconds_batch": 18.06453359290026, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6227481365203857, "target_z": 1.4687500324239748, "ce": 5.982900142669678, "genuineness": 4.582659721374512, "identity": 0.6635078191757202, "hypothesis": "Alguien ha movido la pequeña silla azul. Estaba aquí, junto a la estantería alta. La colocaremos después de comer.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e684089122770917eb486f01ea45421839eafa6eeacf493f6f0682504ad528b8.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "e688cdf0b5e3f37d750e2ddf08ab5c0d0e31776f66db373eae41d003030b2f77", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d6711efbf21336aed12cc4ab7808f4ed9839874b335aaf281b5e7013dd4628c6", "generation_seconds_batch": 16.797331872046925, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5258069038391113, "target_z": 1.398437389218086, "ce": 5.863759994506836, "genuineness": 4.193545341491699, "identity": 0.3523275852203369, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e688cdf0b5e3f37d750e2ddf08ab5c0d0e31776f66db373eae41d003030b2f77.mp3" }, { "id": "heldout-en-3", "language": "en", "sentences": [ "Someone moved the little blue chair.", "It used to stand here... beside the tall bookshelf.", "We can put it back after lunch." ], "text": "Someone moved the little blue chair. It used to stand here... beside the tall bookshelf. We can put it back after lunch.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-3", "key": "e6f1e80cc6606ef00fdc322d5829b94fd65ad80a74ab015f88321e5dc194804c", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Someone moved the little blue chair.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can put it back after lunch.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Someone moved the little blue chair.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can put it back after lunch.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "72819531df7b5bd61590372fa5401906829091ee734254659ee86c373dffd64d", "generation_seconds_batch": 20.027848617057316, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6152608394622803, "target_z": 0.012695326347739263, "ce": 6.459460258483887, "genuineness": 4.331617832183838, "identity": 0.2704594135284424, "hypothesis": "Someone moved the little blue chair. It used to stand here, beside the tall bookshelf. We can put it back after lunch.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e6f1e80cc6606ef00fdc322d5829b94fd65ad80a74ab015f88321e5dc194804c.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "e736526479b56e75f6382a27956e224bc7356a3917942b79dd2b826a784f6044", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "382284101527d878a4a05f2053c2ca69fae3efe0a46c143126e00e3cbc721824", "generation_seconds_batch": 22.932823816896416, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4415748119354248, "target_z": -0.11328124054300734, "ce": 6.173524379730225, "genuineness": 4.469690799713135, "identity": 0.8374302387237549, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e736526479b56e75f6382a27956e224bc7356a3917942b79dd2b826a784f6044.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "e7d97e17ab7c46b48e84259c43bd2569c1f30ee4966a4144e35dd398af020502", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e891a4e8b720448144bf69ff9e1a4b82165c6260bf0dcbe55ca8f2f26157d41f", "generation_seconds_batch": 22.968272380996495, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.055534601211548, "target_z": 0.33203118650304925, "ce": 6.1020402908325195, "genuineness": 4.325342178344727, "identity": 0.5445438623428345, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e7d97e17ab7c46b48e84259c43bd2569c1f30ee4966a4144e35dd398af020502.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "e7de97041864208c24f4540322b7fe5f49f91ffa0fe29266f2d81b4a7400a302", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1fc05c76f4faba23ca160d04b4505011891dfd5752e891717fee3665cdb8c3ca", "generation_seconds_batch": 15.358593748998828, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.859714984893799, "target_z": 1.640624940556046, "ce": 5.677107334136963, "genuineness": 4.432034492492676, "identity": 0.30407989025115967, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e7de97041864208c24f4540322b7fe5f49f91ffa0fe29266f2d81b4a7400a302.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "e8169814f527027fd84d0381996764c81b66c156ceddf8c0b9b30c87774eede9", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "66bed515b3d267b458ae5c978c3bd5c2b89d37456407fe0312f5dc121c862ed8", "generation_seconds_batch": 23.315708325011656, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5042645931243896, "target_z": 1.3828124918940063, "ce": 5.824047088623047, "genuineness": 4.243753433227539, "identity": 0.590313196182251, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschalten.", "wer": 0.0625, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e8169814f527027fd84d0381996764c81b66c156ceddf8c0b9b30c87774eede9.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "e8a05df4526ac34e65ac19e08e6a8004569a89d2223f3b5ef5ac07822d1ec946", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d66af2dd5d7d7bb9a7537803c9771b56ed97b73b98ecae295ce4358547b709cb", "generation_seconds_batch": 23.943483604001813, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.719689130783081, "target_z": 1.539062502701998, "ce": 6.046441555023193, "genuineness": 4.356721878051758, "identity": 0.6606800556182861, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e8a05df4526ac34e65ac19e08e6a8004569a89d2223f3b5ef5ac07822d1ec946.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "e8b3cc9e6c526ca06d190336deff3067bd3960fd655e2e1dae774918c3190739", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1878819fd49173351976a966c26da2750a5132d29e5461c0317b2615eeaab078", "generation_seconds_batch": 24.073619749979116, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3319249153137207, "target_z": 1.25781244866204, "ce": 5.704906940460205, "genuineness": 4.432034492492676, "identity": 0.6123198866844177, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e8b3cc9e6c526ca06d190336deff3067bd3960fd655e2e1dae774918c3190739.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "e92a443492a124286095a56e2975aa28ebc068f252bfd0a3841a33bb8e05558f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "01acd31dfaa3882f80f95976ea0f9d108b24e25c3312c35a7b141ba9bccc04c5", "generation_seconds_batch": 23.054025576915592, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.527531147003174, "target_z": 2.1250000432319665, "ce": 5.879645347595215, "genuineness": 4.507347106933594, "identity": 0.6423183679580688, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e92a443492a124286095a56e2975aa28ebc068f252bfd0a3841a33bb8e05558f.mp3" }, { "id": "heldout-de-0", "language": "de", "sentences": [ "Deine Nachricht kam, als ich schon weg war.", "Ich las sie zweimal... dann legte ich das Handy weg.", "Sag mir, wann wir reden können." ], "text": "Deine Nachricht kam, als ich schon weg war. Ich las sie zweimal... dann legte ich das Handy weg. Sag mir, wann wir reden können.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-0", "key": "e944e67db5d24a9ec896c39792d7f7578c1cb2b30c7314d493dfa4fefc12097a", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wann wir reden können.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Natural clear speech.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wann wir reden können.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "5ac42fcbce5235f250ccd412f584e755f5f3b040243df8ab107b5f40cefd5332", "generation_seconds_batch": 15.556458711973391, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.687375545501709, "target_z": 1.5156250702519456, "ce": 5.760505676269531, "genuineness": 4.632867813110352, "identity": 0.35834330320358276, "hypothesis": "Deine Nachricht kam, als ich schon weg war. Ich las sie zweimal, dann legte ich das Handy weg. Sag mir, wann wir reden können.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e944e67db5d24a9ec896c39792d7f7578c1cb2b30c7314d493dfa4fefc12097a.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "e977533e3339856ce4d75d362ee2479db6e98fec1b1ab5c7704093de210ecaa8", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "88a9c4bdb5e04f038132b9a7a8b32fb043d4aedf99f5c45ce7ed50df0af965d4", "generation_seconds_batch": 24.543410731013864, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5365781784057617, "target_z": 1.4062499243440587, "ce": 5.297844886779785, "genuineness": 4.250029563903809, "identity": 0.6465438604354858, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás lista.", "wer": 0.06666666666666667, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e977533e3339856ce4d75d362ee2479db6e98fec1b1ab5c7704093de210ecaa8.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "e97bcdb879a400e47b0a9429ddb26671a3b61a4ebdc87fc2bea358efd18abbcf", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3318d39f8a1c66cecaff676396dbb14bb6d6eff84a76e2ff60e19f33935ecca8", "generation_seconds_batch": 23.32810746098403, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.735090732574463, "target_z": 0.09960939053648796, "ce": 6.38003396987915, "genuineness": 4.582659721374512, "identity": 0.5744701027870178, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e97bcdb879a400e47b0a9429ddb26671a3b61a4ebdc87fc2bea358efd18abbcf.mp3" }, { "id": "heldout-de-3", "language": "de", "sentences": [ "Jemand hat den kleinen blauen Stuhl umgestellt.", "Er stand hier... neben dem hohen Bücherregal.", "Wir stellen ihn nach dem Essen zurück." ], "text": "Jemand hat den kleinen blauen Stuhl umgestellt. Er stand hier... neben dem hohen Bücherregal. Wir stellen ihn nach dem Essen zurück.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-3", "key": "e9d266c3907611a53d0c5f7688a7e231bba29ead6f87b8ae8d29cb37ab2fae41", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ef2f238896ecf6ac7ceb577679948cf0e786ff9b55119e8bb21207d6598e289f", "generation_seconds_batch": 15.478994615958072, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.30424165725708, "target_z": -0.2128906310794953, "ce": 6.467403411865234, "genuineness": 4.6203155517578125, "identity": 0.5337745547294617, "hypothesis": "Jemand hat den kleinen blauen Stuhl umgestellt. Er stand hier neben dem hohen Bücherregal. Wir stellen ihn nach dem Essen zurück.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/e9d266c3907611a53d0c5f7688a7e231bba29ead6f87b8ae8d29cb37ab2fae41.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "ea2b8a9a91171548a073cb4625779d54abb1208aa5fb4b5521b933e488f5c9d4", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f1bb737404e813588d8779010d737374b8d07613d788e95ceea3c843712c6cab", "generation_seconds_batch": 21.215285453014076, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0484249591827393, "target_z": -0.39843756214595183, "ce": 6.086154937744141, "genuineness": 4.369274139404297, "identity": 0.5492993593215942, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ea2b8a9a91171548a073cb4625779d54abb1208aa5fb4b5521b933e488f5c9d4.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "ea57fe65fe01192251a5c31842b6584ffcaf55da9806c8b2859997e6bf84be3f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "b12a450512eafc268935c6ba021253264024127f29d216e1335e8b39c415a39f", "generation_seconds_batch": 24.98371786705684, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.441361427307129, "target_z": 2.062500108079916, "ce": 5.43088436126709, "genuineness": 4.607763767242432, "identity": 0.8672847747802734, "hypothesis": "La porte est encore ouverte, j'ai posé ton manteau sur la chaise, dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ea57fe65fe01192251a5c31842b6584ffcaf55da9806c8b2859997e6bf84be3f.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "ea5c2089fe7f146ff4734409a890c1220d605474c459cb853103ebf2737be979", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5b61dbf08eadbadfc904b84d654532d44bde31053b81ec0762ac0d45b06e091e", "generation_seconds_batch": 22.18766916997265, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1272716522216797, "target_z": 1.1093749729800209, "ce": 6.070269584655762, "genuineness": 4.042920112609863, "identity": 0.1634877622127533, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ea5c2089fe7f146ff4734409a890c1220d605474c459cb853103ebf2737be979.mp3" }, { "id": "heldout-es-3", "language": "es", "sentences": [ "Alguien ha movido la pequeña silla azul.", "Estaba aquí... junto a la estantería alta.", "La colocaremos después de comer." ], "text": "Alguien ha movido la pequeña silla azul. Estaba aquí... junto a la estantería alta. La colocaremos después de comer.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-3", "key": "ea8389fe373ec9be7ac7a0bd01a178e17180f829b0083fcee046ff3387b30457", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La colocaremos después de comer.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Natural clear speech.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Natural clear speech.) [3.2 seconds duration] La colocaremos después de comer.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c9d9322a0ca77d307e75abcda6683c38d54da81ce8486d4692fd8d7688cd4038", "generation_seconds_batch": 14.859678464010358, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.256526470184326, "target_z": 1.2031250486359624, "ce": 5.450741291046143, "genuineness": 4.381826400756836, "identity": 0.4785759747028351, "hypothesis": "Alguien ha movido la pequeña silla azul. Estaba aquí, junto a la estantería alta. La colocaremos después de comer.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ea8389fe373ec9be7ac7a0bd01a178e17180f829b0083fcee046ff3387b30457.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "eab4956a939d29ac9a9edc498b74f7183fe84a5c113a544f4981f0e0971e436a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ad4c63bee5a0521c1c09ded2abf24fd9bd87b174e97bd84324a8b726088e48b7", "generation_seconds_batch": 22.23857655900065, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7179651260375977, "target_z": 0.8125000216159832, "ce": 6.340320587158203, "genuineness": 4.457139015197754, "identity": 0.21078293025493622, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/eab4956a939d29ac9a9edc498b74f7183fe84a5c113a544f4981f0e0971e436a.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "eacd9f3aba8b8845a720b2e094871b2667121fb70cdc70303098a30e863286be", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9be2cfd11955a3231632711b4f276b835322c3b46cadbc9a6eec890e5952eb42", "generation_seconds_batch": 23.290174574009143, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0797698497772217, "target_z": 0.34960930407255497, "ce": 6.133810997009277, "genuineness": 4.444586753845215, "identity": 0.7540993094444275, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/eacd9f3aba8b8845a720b2e094871b2667121fb70cdc70303098a30e863286be.mp3" }, { "id": "heldout-en-4", "language": "en", "sentences": [ "I wrote the number on the envelope.", "Read it carefully... and check the last two digits.", "Call me when you have a moment." ], "text": "I wrote the number on the envelope. Read it carefully... and check the last two digits. Call me when you have a moment.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-4", "key": "eb342567765ac90ae8da0d80cb2ca97aefe1a1fa23b9bb5a07d403d9dc7b99a3", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I wrote the number on the envelope.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Call me when you have a moment.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I wrote the number on the envelope.\n(Natural clear speech.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Natural clear speech.) [3.2 seconds duration] Call me when you have a moment.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1acbbd72a323ab3c41b98039f5db4d9dbf82767974ad360b3e779841138509dd", "generation_seconds_batch": 17.297314282972366, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6264095306396484, "target_z": 0.7460936459730807, "ce": 6.054384231567383, "genuineness": 3.9425034523010254, "identity": 0.4529355764389038, "hypothesis": "I wrote the number on the envelope. Read it carefully. And check the last two digits. Call me when you have a moment.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/eb342567765ac90ae8da0d80cb2ca97aefe1a1fa23b9bb5a07d403d9dc7b99a3.mp3" }, { "id": "heldout-en-3", "language": "en", "sentences": [ "Someone moved the little blue chair.", "It used to stand here... beside the tall bookshelf.", "We can put it back after lunch." ], "text": "Someone moved the little blue chair. It used to stand here... beside the tall bookshelf. We can put it back after lunch.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-en-3", "key": "eb68753aa6c8ceb60e39d4d35cbe1bde7506abab466c8e3cadedca24d83841a7", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Someone moved the little blue chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can put it back after lunch.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Someone moved the little blue chair.\n(Natural clear speech.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Natural clear speech.) [3.2 seconds duration] We can put it back after lunch.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8773e74f1c8dd59bde9ace2c27964a6cf07f960ac9423b5dd1a3043dbf2566a5", "generation_seconds_batch": 17.178146843914874, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.450408458709717, "target_z": 1.3437499891920084, "ce": 6.062326908111572, "genuineness": 4.331617832183838, "identity": 0.4210236072540283, "hypothesis": "Someone moved the little blue chair. It used to stand here beside the tall bookshelf. We can put it back after lunch.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/eb68753aa6c8ceb60e39d4d35cbe1bde7506abab466c8e3cadedca24d83841a7.mp3" }, { "id": "heldout-fr-4", "language": "fr", "sentences": [ "J’ai écrit le numéro sur l’enveloppe.", "Lis-le bien... et vérifie les deux derniers chiffres.", "Appelle-moi quand tu auras un moment." ], "text": "J’ai écrit le numéro sur l’enveloppe. Lis-le bien... et vérifie les deux derniers chiffres. Appelle-moi quand tu auras un moment.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-4", "key": "eba092aa3c1803d7c77d24ea89146bec9bac16f65165073cfecda99f3037afbb", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Natural clear speech.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Natural clear speech.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "111eff1ce6c65e80673a2f146588a04067e81d0a130c9ef1987da9417a1df7f2", "generation_seconds_batch": 24.299769907956943, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.021283149719238, "target_z": 1.7578122757341739, "ce": 5.530168056488037, "genuineness": 4.432034492492676, "identity": 0.8049426078796387, "hypothesis": "J'ai écrit le numéro sur l'enveloppe, lis le bien et vérifie les deux derniers chiffres. Appelle-moi quand tu auras un moment.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/eba092aa3c1803d7c77d24ea89146bec9bac16f65165073cfecda99f3037afbb.mp3" }, { "id": "heldout-fr-2", "language": "fr", "sentences": [ "La route tourne à gauche après le pont.", "Continue... jusqu’à la petite barrière en bois.", "Je t’attendrai de l’autre côté." ], "text": "La route tourne à gauche après le pont. Continue... jusqu’à la petite barrière en bois. Je t’attendrai de l’autre côté.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-2", "key": "ebe89981a89e9506d02bacdfa583f9bfe1f04faa699193749ab6419de85c588f", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "728359eb4836c2bd5e07a2e078adcaafa28412eaee5798ddaacf2f5226a306ae", "generation_seconds_batch": 16.59471099707298, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.319429874420166, "target_z": 0.5234376053779183, "ce": 6.078212261199951, "genuineness": 4.419482707977295, "identity": 0.7622629404067993, "hypothesis": "La route tourne à gauche après le pont. Continue jusqu'à la petite barrière en bois. Je t'attendrai de l'autre côté.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ebe89981a89e9506d02bacdfa583f9bfe1f04faa699193749ab6419de85c588f.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "ec8b3ef51406ee2b7d5b419505487709c027d6aeb51ad72a163a46446ba263f7", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6901a104de4c5fed14f3a15b63d3afdbae3e8c11036ad14ec896e85c53b262cf", "generation_seconds_batch": 26.81906083296053, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.245755195617676, "target_z": 1.1953125135099896, "ce": 6.046441555023193, "genuineness": 4.344170093536377, "identity": 0.6826578378677368, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ec8b3ef51406ee2b7d5b419505487709c027d6aeb51ad72a163a46446ba263f7.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "ec9d1ebf2b20612f5ab26e1910644421edcc5dbc76faf0a592f2137f26675a1a", "bucket": "vn_S_ASMR_reg", "phase": "wording", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "46084190b573620bb1d9b5f031b93b9c659c2ae9723c84e02c8298fbad2c5518", "generation_seconds_batch": 16.535874380962923, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5671679973602295, "target_z": 0.7031250486359623, "ce": 6.221180438995361, "genuineness": 4.13078498840332, "identity": 0.5307742357254028, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ec9d1ebf2b20612f5ab26e1910644421edcc5dbc76faf0a592f2137f26675a1a.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "ecf5db13245f3498740f0b7ada96eabd64c5c02cad794f86ba8d5cf5849f387e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "abe678d4a96c1765a6c495aa764aa2bd9cd123830873f348bc1fb139cc99b284", "generation_seconds_batch": 24.186401372076944, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4540700912475586, "target_z": 0.62109377566898, "ce": 6.062326908111572, "genuineness": 4.557555198669434, "identity": 0.6870741844177246, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ecf5db13245f3498740f0b7ada96eabd64c5c02cad794f86ba8d5cf5849f387e.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "ed4843b75b1c9af6f428dc6fd0fbe30ee117f49e351772d613cbf27e5b272c44", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e7a87153966c06c1e0ffc62cc739accb9b6e3f61f4bfdbb6eeeed9231449d60a", "generation_seconds_batch": 17.46393624099437, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2996113300323486, "target_z": 1.2343750162119875, "ce": 6.260893821716309, "genuineness": 4.18099308013916, "identity": 0.25549864768981934, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ed4843b75b1c9af6f428dc6fd0fbe30ee117f49e351772d613cbf27e5b272c44.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "ed788622deb1c032deca85915431291cdb25c47b8d44c2a3d5b8301719e1bc22", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9e96e01a205cb788281a11a89a85a03c5dc2878d4803f5a9650d0064d388718c", "generation_seconds_batch": 23.071059967041947, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.838172674179077, "target_z": 1.6250000432319665, "ce": 5.9431867599487305, "genuineness": 4.7332844734191895, "identity": 0.5668792128562927, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ed788622deb1c032deca85915431291cdb25c47b8d44c2a3d5b8301719e1bc22.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "edbfce715e13816c556acb94a02f9927dd239670d63e5cab00e193f76546b8d7", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "73b0e3a4a7ee80fa8a2d31deda750069261396e6e74a659e8184ead7c6c16cbe", "generation_seconds_batch": 22.459264148026705, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.577939033508301, "target_z": 0.7109374108340691, "ce": 6.419747352600098, "genuineness": 4.607763767242432, "identity": 0.7083075046539307, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/edbfce715e13816c556acb94a02f9927dd239670d63e5cab00e193f76546b8d7.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "ee3ee554edd82f6d08cb6c5177309a4eafe94b34c3dd4045c3faa45e73d6bf83", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8e13a886a7dacf98810cd0ebe54dfafba3c35b6dc6ee4a8e5c6a7b6163cc40c0", "generation_seconds_batch": 16.83830051100813, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2763447761535645, "target_z": 0.49218746487402726, "ce": 6.308549880981445, "genuineness": 4.457139015197754, "identity": 0.7033502459526062, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ee3ee554edd82f6d08cb6c5177309a4eafe94b34c3dd4045c3faa45e73d6bf83.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "ee9c7b22107c7359eb5659864b930de74f28cb359591bb5f24fc59f55d8d7374", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "530b5ac0a2577aa1fe28c223074195382780eabac876ea5d50865deffea627fd", "generation_seconds_batch": 23.488602872006595, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8310627937316895, "target_z": 0.8945311216550995, "ce": 6.260893821716309, "genuineness": 4.607763767242432, "identity": 0.6482038497924805, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ee9c7b22107c7359eb5659864b930de74f28cb359591bb5f24fc59f55d8d7374.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "eec2dbb586d823778efcf521ba758fc1f25ae26114392aaa906f981aaec91b37", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c1c38ae44d43d3d16ab6a0e2de79981fb127c99db962f310aad64cdbb3a687f7", "generation_seconds_batch": 23.98545291798655, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2709591388702393, "target_z": 0.48828119731104086, "ce": 6.300607204437256, "genuineness": 4.482243061065674, "identity": 0.6059811115264893, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/eec2dbb586d823778efcf521ba758fc1f25ae26114392aaa906f981aaec91b37.mp3" }, { "id": "heldout-es-1", "language": "es", "sentences": [ "Las llaves están en la mesa de la cocina.", "Miré en el cajón... y luego junto a la ventana.", "Llévatelas cuando te vayas." ], "text": "Las llaves están en la mesa de la cocina. Miré en el cajón... y luego junto a la ventana. Llévatelas cuando te vayas.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-es-1", "key": "ef718cd427aabdb7039b57d0d2cb01aba2faf9e9001518c1d4552d5df2fb2f60", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Llévatelas cuando te vayas.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Natural clear speech.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Natural clear speech.) [3.2 seconds duration] Llévatelas cuando te vayas.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "76a15220189d70d3eec669e78406bf357bcb62f170c45aee1ef724d7be8ecf84", "generation_seconds_batch": 24.245687233982608, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.450408458709717, "target_z": 1.3437499891920084, "ce": 5.5738525390625, "genuineness": 4.218649387359619, "identity": 0.5059900283813477, "hypothesis": "Las llaves están en la mesa de la cocina. Miren el cajón y luego, junto a la ventana, llévatelas cuando te vayas.", "wer": 0.08695652173913043, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ef718cd427aabdb7039b57d0d2cb01aba2faf9e9001518c1d4552d5df2fb2f60.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "efa5cd221f026564881074ba7efd934060c5f86e607aa7a8c37f12848ee2f5fc", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8922362505b7f47a98e87ef5415670cb69a7bb2f2230969ecbc85b2a3cbe1945", "generation_seconds_batch": 25.54708585201297, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.247479438781738, "target_z": 1.92187516752387, "ce": 5.5539960861206055, "genuineness": 4.507347106933594, "identity": 0.5882800817489624, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/efa5cd221f026564881074ba7efd934060c5f86e607aa7a8c37f12848ee2f5fc.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "efa8fcfec2576700e683a4dd55aaca2c3e06882721f4c0850eb7a87458a34fdf", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6e4ba040e783eafac21558b88a9831418e23abeb3c2b589b6b4e43e4e9574bbb", "generation_seconds_batch": 24.548229896929115, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4540700912475586, "target_z": 0.62109377566898, "ce": 5.95112943649292, "genuineness": 4.15588903427124, "identity": 0.4082757234573364, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/efa8fcfec2576700e683a4dd55aaca2c3e06882721f4c0850eb7a87458a34fdf.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "efa92fc8fd8470fa45fa0cc5e65819155c485255a9a75a12fb937be62ab489ab", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "965e9bb084b2b1a7b0b32dc28024899969999b062d74a9e334c1b18522fd1459", "generation_seconds_batch": 25.89438219089061, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8812575340270996, "target_z": 1.6562500108079916, "ce": 6.070269584655762, "genuineness": 4.65797233581543, "identity": 0.6032178401947021, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/efa92fc8fd8470fa45fa0cc5e65819155c485255a9a75a12fb937be62ab489ab.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "efb0c0ddacdc35f9c04c40ead986e3c34323dca623f8c2f94a9f5f980e286242", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8b559123e484fa8d0a30068a2d27575de97979216f0c4fb4315f7e9de2375f3c", "generation_seconds_batch": 25.0993486889638, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.191899061203003, "target_z": 1.1562500108079916, "ce": 6.133810997009277, "genuineness": 4.595211505889893, "identity": 0.7999258041381836, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/efb0c0ddacdc35f9c04c40ead986e3c34323dca623f8c2f94a9f5f980e286242.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "f089846ac8fcab7dfb2cedc8b9be5c3122036154052f8e401fa3eec431628830", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e2b865576daed9d3eb5197c248356ee72323f1def873363a988e91825dcc5b56", "generation_seconds_batch": 24.99255053792149, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7950878143310547, "target_z": 1.5937500756559413, "ce": 6.1894097328186035, "genuineness": 4.400654315948486, "identity": 0.7786912322044373, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f089846ac8fcab7dfb2cedc8b9be5c3122036154052f8e401fa3eec431628830.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "f0db0ed531849083624549d67c7c0561f11c8f98fbb0fc1fe4ce7fa7e4368f71", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d0d6422de6992572845147c82a8a18d016070da09eaafe62709399db560a1757", "generation_seconds_batch": 17.40729280503001, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.417339563369751, "target_z": -0.1308593581125131, "ce": 6.046441555023193, "genuineness": 4.419482707977295, "identity": 0.5497045516967773, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f0db0ed531849083624549d67c7c0561f11c8f98fbb0fc1fe4ce7fa7e4368f71.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "f186624c1315a7c76f5523a153abc3d6d9e098694a6276416ca100cdd76c4593", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a082e4ea542093e78376fa205c8dafd7e0c5daefb86b56c2ff476833ed1e3148", "generation_seconds_batch": 24.249534575035796, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.592158317565918, "target_z": 2.171874908132071, "ce": 5.681078910827637, "genuineness": 4.557555198669434, "identity": 0.7830448746681213, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f186624c1315a7c76f5523a153abc3d6d9e098694a6276416ca100cdd76c4593.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "f1e891c73b8adff8215662fb4031dc3cd16c3dfdc1dafafcfabb7981ec08e89b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "223a2efd7d2523741039cab63413993ad3bdb8001bbbcc985966fba034b9ec97", "generation_seconds_batch": 22.431560627068393, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8274013996124268, "target_z": 1.6171875081059937, "ce": 5.677107334136963, "genuineness": 4.570107460021973, "identity": 0.4418342113494873, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f1e891c73b8adff8215662fb4031dc3cd16c3dfdc1dafafcfabb7981ec08e89b.mp3" }, { "id": "heldout-fr-0", "language": "fr", "sentences": [ "Ton message est arrivé après mon départ.", "Je l’ai relu... puis j’ai rangé mon téléphone.", "Dis-moi quand nous pourrons parler." ], "text": "Ton message est arrivé après mon départ. Je l’ai relu... puis j’ai rangé mon téléphone. Dis-moi quand nous pourrons parler.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-0", "key": "f21bc0adf2df6aeda4aa5bc68657c60b57559ffed3b647118e970ebfd0ba6ad8", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7d92ec9886822e47d0db11be498e535621cb78912f74f666e00063a14f3b880f", "generation_seconds_batch": 15.886211984907277, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1390116214752197, "target_z": 0.3925780743375393, "ce": 6.173524379730225, "genuineness": 4.507347106933594, "identity": 0.7314589619636536, "hypothesis": "Ton message est arrivé après mon départ, je l'ai relu, puis j'ai rangé mon téléphone. Dis-moi quand nous pourrons parler.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f21bc0adf2df6aeda4aa5bc68657c60b57559ffed3b647118e970ebfd0ba6ad8.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "f28f34ca38cf93756de9da817eab99660315af4359db0397dcd433262625103f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "d4888a7d88eebd8dc6c210ee0795fb7510a8a1ec90d2f26e7942c0323ca0e146", "generation_seconds_batch": 22.03802254400216, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.245755195617676, "target_z": 1.1953125135099896, "ce": 6.197352409362793, "genuineness": 4.381826400756836, "identity": 0.7256649732589722, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f28f34ca38cf93756de9da817eab99660315af4359db0397dcd433262625103f.mp3" }, { "id": "heldout-nl-0", "language": "nl", "sentences": [ "Je bericht kwam nadat ik vertrokken was.", "Ik las het twee keer... en legde mijn telefoon weg.", "Zeg maar wanneer we kunnen praten." ], "text": "Je bericht kwam nadat ik vertrokken was. Ik las het twee keer... en legde mijn telefoon weg. Zeg maar wanneer we kunnen praten.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-0", "key": "f2f90fa0093ae5f499cbd33aabbf1d9e99cc57937873bd64f9c07d5048be4559", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Natural clear speech.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c2587689de6823b6b2ea3a213634c31fff64da3ff6cad55cb0dea78945d6fac4", "generation_seconds_batch": 15.034878571983427, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.84183406829834, "target_z": 0.9023436567810723, "ce": 5.478540420532227, "genuineness": 4.231201171875, "identity": 0.23987700045108795, "hypothesis": "Je bericht kwam nadat ik vertrokken was. Ik las het twee keer en legde mijn telefoon weg. Zeg maar wanneer we kunnen praten.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f2f90fa0093ae5f499cbd33aabbf1d9e99cc57937873bd64f9c07d5048be4559.mp3" }, { "id": "heldout-fr-0", "language": "fr", "sentences": [ "Ton message est arrivé après mon départ.", "Je l’ai relu... puis j’ai rangé mon téléphone.", "Dis-moi quand nous pourrons parler." ], "text": "Ton message est arrivé après mon départ. Je l’ai relu... puis j’ai rangé mon téléphone. Dis-moi quand nous pourrons parler.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-0", "key": "f386ac49a2a29fd861c5a9b7543901aef2e2d3714d311210ca39dcb665f458bc", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Natural clear speech.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "b4fc69a074d49580c8e20b3319f4c0195a54924f835015dd91db9071e1975d30", "generation_seconds_batch": 23.048492497066036, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.053597450256348, "target_z": 1.781250226967824, "ce": 5.7128496170043945, "genuineness": 4.3127899169921875, "identity": 0.6345981359481812, "hypothesis": "Ton message est arrivé après mon départ, je l'ai relu, puis j'ai rangé mon téléphone. Dis-moi quand nous pourrons parler.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f386ac49a2a29fd861c5a9b7543901aef2e2d3714d311210ca39dcb665f458bc.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "f3d20c7077adc446cef11d77839c1dd10646a4a20c96e69417f1cc90f0148704", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4bb51b9d78ef699767a627d7d0b4d40ccaa11ddb27ba83869564f8a11bbed434", "generation_seconds_batch": 23.223652250948362, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.301335334777832, "target_z": 1.960937497298002, "ce": 5.919358730316162, "genuineness": 4.607763767242432, "identity": 0.6970577239990234, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f3d20c7077adc446cef11d77839c1dd10646a4a20c96e69417f1cc90f0148704.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "f3d380d8c065cc51efe8035476bafe7f3dcac27e9a9ae7b4e94b54e279abb6ae", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "39c291b3fa77ab49924b3ea258386367ed8df3f330d19e4e9f7a83cdcfeab9c2", "generation_seconds_batch": 22.03919477108866, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2888400554656982, "target_z": 1.2265624810860147, "ce": 6.117925643920898, "genuineness": 4.444586753845215, "identity": 0.6807153224945068, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f3d380d8c065cc51efe8035476bafe7f3dcac27e9a9ae7b4e94b54e279abb6ae.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "f3f6c1a6ad14371ebea63782ebefc1e53861490b227494618085aa6b46c67bb7", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2c69893d3a2c8a69027dcabaf401a6bc7d9273f0df79627e34667f273734daa5", "generation_seconds_batch": 25.293777625076473, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9764745235443115, "target_z": 1.0, "ce": 5.49045467376709, "genuineness": 4.030367851257324, "identity": 0.5160014033317566, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás libre.", "wer": 0.06666666666666667, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f3f6c1a6ad14371ebea63782ebefc1e53861490b227494618085aa6b46c67bb7.mp3" }, { "id": "heldout-nl-2", "language": "nl", "sentences": [ "De weg gaat linksaf na de brug.", "Loop verder... tot aan het kleine houten hek.", "Ik wacht aan de andere kant." ], "text": "De weg gaat linksaf na de brug. Loop verder... tot aan het kleine houten hek. Ik wacht aan de andere kant.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-nl-2", "key": "f444758d965a0849b558da26fbc0c5fa2691b35f0b74650def630e5642f16a32", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik wacht aan de andere kant.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Natural clear speech.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Natural clear speech.) [3.2 seconds duration] Ik wacht aan de andere kant.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d91f99640c36beec2b566a6d96215be72ce08fdd6469f1f934a16d25c1031666", "generation_seconds_batch": 17.09770359203685, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2134413719177246, "target_z": 1.1718749081320712, "ce": 6.173524379730225, "genuineness": 4.557555198669434, "identity": 0.33896082639694214, "hypothesis": "De weg gaat linksaf naar de brug. Loop verder, tot aan het kleine houten hek. Ik wacht aan de andere kant.", "wer": 0.047619047619047616, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f444758d965a0849b558da26fbc0c5fa2691b35f0b74650def630e5642f16a32.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "f45b386a32b2d9b15b541e537bf58a4c64252d9aeeed9dad8a3860caf4081f3b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "25fa79f18ab6dabfa7eaaee2ee41b3c3e8e003c2f4ebb8ad90bbf3c7d4d81c36", "generation_seconds_batch": 23.905337012954988, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.051873207092285, "target_z": 1.0546875729539433, "ce": 6.221180438995361, "genuineness": 4.545003414154053, "identity": 0.6965397596359253, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f45b386a32b2d9b15b541e537bf58a4c64252d9aeeed9dad8a3860caf4081f3b.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "f49d7a9c06da379610ab25db4fb1d40832cb5c3c59b330a1e4dc0131039ae190", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "25bb94fafa27e6b8e90d71cfcfaf91026b0dbc9e93952a1e5a08b80abea0af10", "generation_seconds_batch": 23.876749251037836, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.07513952255249, "target_z": 1.7968749513640376, "ce": 6.125868320465088, "genuineness": 4.6956281661987305, "identity": 0.8228939771652222, "hypothesis": "J'ai trouvé la petite boîte, elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f49d7a9c06da379610ab25db4fb1d40832cb5c3c59b330a1e4dc0131039ae190.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "f516ac377b9faff1d5698299f2e570c8ddf7771e7087ecf1412274a7cdea17c3", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "641b7d79cabf83b00392d49fb9356be2b44ca9b4005a05d5b08120384ba950e2", "generation_seconds_batch": 23.202125952928327, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.470226764678955, "target_z": 0.6328124054300733, "ce": 6.157639026641846, "genuineness": 4.7458367347717285, "identity": 0.5758295655250549, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f516ac377b9faff1d5698299f2e570c8ddf7771e7087ecf1412274a7cdea17c3.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "f52a7e09abf81211d7a51f3bd259f99430be17ee135e7722ffe8256101936abc", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e657f767ca0aca12a325c1fc3ba331330bc82d6deff831587350a5bed6fef90e", "generation_seconds_batch": 22.18766916997265, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0087881088256836, "target_z": 1.0234374324500524, "ce": 6.300607204437256, "genuineness": 4.281409740447998, "identity": 0.17975816130638123, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f52a7e09abf81211d7a51f3bd259f99430be17ee135e7722ffe8256101936abc.mp3" }, { "id": "heldout-fr-4", "language": "fr", "sentences": [ "J’ai écrit le numéro sur l’enveloppe.", "Lis-le bien... et vérifie les deux derniers chiffres.", "Appelle-moi quand tu auras un moment." ], "text": "J’ai écrit le numéro sur l’enveloppe. Lis-le bien... et vérifie les deux derniers chiffres. Appelle-moi quand tu auras un moment.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-4", "key": "f5335cca14ddee98ce57cedf235b84b853fff80601dd892d646af1ae73bcc693", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Natural clear speech.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Natural clear speech.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c9ec5c3c3fca70a2fed98202bdac4e4ff65811cf5ebc2ed8c2d7e0c3cb395fe4", "generation_seconds_batch": 24.299769907956943, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.107453346252441, "target_z": 1.820312556741956, "ce": 5.383228778839111, "genuineness": 4.331617832183838, "identity": 0.7441379427909851, "hypothesis": "J'ai écrit le numéro sur l'enveloppe, lis-le bien et vérifiez les deux derniers chiffres. Appelle-moi quand tu auras un moment.", "wer": 0.045454545454545456, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f5335cca14ddee98ce57cedf235b84b853fff80601dd892d646af1ae73bcc693.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "f67eb583c97aa9d964c444c517174813d0dcbd22009780599de874984a31e9c0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "42bbc0cb895491c606d5f05b7744fdbd599dfa4ba82c789ee323facb20866f98", "generation_seconds_batch": 17.33628339599818, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.960317611694336, "target_z": 0.9882811973110408, "ce": 5.871703147888184, "genuineness": 4.086852550506592, "identity": 0.4486689269542694, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f67eb583c97aa9d964c444c517174813d0dcbd22009780599de874984a31e9c0.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "f6d7d4382b19078ad9382a243fd14bd1176ff236326549ad500994dee30c76ff", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 131, "seconds_generated": 10.48, "mp3_sha256": "e77e3137f3acf1a7c05691acfd9b738603572bc6769967ce48226298956eb30d", "generation_seconds_batch": 23.77853731403593, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.51503586769104, "target_z": 1.3906250270199791, "ce": 6.038498878479004, "genuineness": 4.168440818786621, "identity": 0.3427186608314514, "hypothesis": "I found the little box. It was beside the old window. We can look inside.", "wer": 0.0625, "duration_error_s": 0.28000000000000114, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f6d7d4382b19078ad9382a243fd14bd1176ff236326549ad500994dee30c76ff.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "f6fd775c73d7ead2f6bb803bf4dad1fc838746dba79cd3e4f4238893764b786b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d094d51d2e54dfaeeea542059386eeb09789e7f33c74b89473a925cd263131be", "generation_seconds_batch": 17.535287783946842, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9909073114395142, "target_z": 0.2851562351390115, "ce": 6.109982967376709, "genuineness": 4.406930446624756, "identity": 0.6461523771286011, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f6fd775c73d7ead2f6bb803bf4dad1fc838746dba79cd3e4f4238893764b786b.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "f7159b1bcd0ca9886510ff4fc813be5e8cf67af8744c24d7657adf615da68036", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f8579c8fa9d27ddf5259c0b88daa7c9b5cdb0ec79e16b8286d664afc53d2c1a2", "generation_seconds_batch": 23.478353463113308, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4954309463500977, "target_z": -0.07421873784100942, "ce": 6.324435234069824, "genuineness": 4.65797233581543, "identity": 0.6343194842338562, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f7159b1bcd0ca9886510ff4fc813be5e8cf67af8744c24d7657adf615da68036.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "f717a246ca10bb908aebd799e47f8de6b4ae14b49ec0a23317205ef5231dee7b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2549478e3c25b410542685051d178518c5d92c49a4a2e8934e61c6bb62bfdb29", "generation_seconds_batch": 32.573457063990645, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6210241317749023, "target_z": 0.7421875513379602, "ce": 6.038498878479004, "genuineness": 4.557555198669434, "identity": 0.6810423135757446, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f717a246ca10bb908aebd799e47f8de6b4ae14b49ec0a23317205ef5231dee7b.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "f76da23fb5cd50a945f208bc89e41e04b369801d2df655198509dbd2726ea83b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b7934ccbc4805c623db930f2d8b46a1d05b7d6fb114ac29c357f91ee49a8f76c", "generation_seconds_batch": 22.957556341891177, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7445155382156372, "target_z": 0.10644531553974765, "ce": 6.3641486167907715, "genuineness": 4.607763767242432, "identity": 0.43238505721092224, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f76da23fb5cd50a945f208bc89e41e04b369801d2df655198509dbd2726ea83b.mp3" }, { "id": "heldout-fr-1", "language": "fr", "sentences": [ "Les clés sont sur la table de la cuisine.", "J’ai regardé dans le tiroir... puis près de la fenêtre.", "Prends-les quand tu partiras." ], "text": "Les clés sont sur la table de la cuisine. J’ai regardé dans le tiroir... puis près de la fenêtre. Prends-les quand tu partiras.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-fr-1", "key": "f7a65885d0e107603bdce9d92b5fc3b68a68524db7b0b06f79b6538794e8bf5c", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Prends-les quand tu partiras.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Prends-les quand tu partiras.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "5347fd16bddc390127e4e2cc7e2dd87341196e53a75e4b6447e0c47d10af694b", "generation_seconds_batch": 16.290645889006555, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7323979139328003, "target_z": 0.09765625675499476, "ce": 6.23706579208374, "genuineness": 4.7081804275512695, "identity": 0.681900680065155, "hypothesis": "Les clés sont sur la table de la cuisine. J'ai regardé dans le tiroir, puis près de la fenêtre. Prends-les quand tu partiras.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f7a65885d0e107603bdce9d92b5fc3b68a68524db7b0b06f79b6538794e8bf5c.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "f81bbd37ec1322ffef503698410189cf3448cbe1eb3ba4f9a0fea778ed70b298", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "135f4e4e23d1ae6a66ae014e365dac77548895c524f716de9d4f23d345e136cd", "generation_seconds_batch": 25.695698769995943, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.556396722793579, "target_z": 0.6953125135099896, "ce": 6.229123115539551, "genuineness": 4.425758361816406, "identity": 0.4952654242515564, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f81bbd37ec1322ffef503698410189cf3448cbe1eb3ba4f9a0fea778ed70b298.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "f865662723b900b9e7cc4e570bb7ba63f9d97c527d57fd4287153e50eebd800f", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f3aaa1200ad9de77c3807576d4cbeb6ab9b5f5cba30e511fcf67b53977e98c5d", "generation_seconds_batch": 23.10274878994096, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7304604053497314, "target_z": 1.5468750378279708, "ce": 5.911416053771973, "genuineness": 4.68307638168335, "identity": 0.7473714351654053, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f865662723b900b9e7cc4e570bb7ba63f9d97c527d57fd4287153e50eebd800f.mp3" }, { "id": "heldout-de-2", "language": "de", "sentences": [ "Die Straße führt hinter der Brücke nach links.", "Geh weiter... bis zu dem kleinen Holztor.", "Ich warte auf der anderen Seite." ], "text": "Die Straße führt hinter der Brücke nach links. Geh weiter... bis zu dem kleinen Holztor. Ich warte auf der anderen Seite.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-2", "key": "f86feddc7af334e5c3d6691e715ece3168f8e7618bd3b63ca7d0171f412fdf9d", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "48635c5a8ca70794e89fb9d9739aec819d57b0bd4ee30b46be3e05197a51f67e", "generation_seconds_batch": 15.55431785795372, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5348541736602783, "target_z": 0.679687443258044, "ce": 6.348263263702393, "genuineness": 4.507347106933594, "identity": 0.6422327160835266, "hypothesis": "Die Straße führt hinter der Brücke nach links. Geh weiter bis zum kleinen Holztor. Ich warte auf der anderen Seite.", "wer": 0.09523809523809523, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f86feddc7af334e5c3d6691e715ece3168f8e7618bd3b63ca7d0171f412fdf9d.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "f872cd9b3e8151aa1b886f82cf0eef0dee5bf7eff60131d16fc34ffc73fe4033", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e4972bd5d2ecf0f742f84df1f26b669935caf066165d97fef4ad5dfdc99f3b4e", "generation_seconds_batch": 20.402951620984823, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.224212646484375, "target_z": 1.179687443258044, "ce": 6.062326908111572, "genuineness": 4.356721878051758, "identity": 0.6167676448822021, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinspringen.", "wer": 0.0625, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f872cd9b3e8151aa1b886f82cf0eef0dee5bf7eff60131d16fc34ffc73fe4033.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "f896a9c9a291f9dc728f93482cbc2a6cccbd8e85545c488576ce3284446401ba", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "14a5a1684eff92c45f32fce33e03fd93ab83ce9ae197c29df440e8263a614cdb", "generation_seconds_batch": 25.028561326907948, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8812575340270996, "target_z": 1.6562500108079916, "ce": 5.919358730316162, "genuineness": 4.482243061065674, "identity": 0.47538459300994873, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f896a9c9a291f9dc728f93482cbc2a6cccbd8e85545c488576ce3284446401ba.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "f8c45be98be65ea1c70e545ecaa6030fd226ae49f79320ec08728f66dcba0355", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5bd7a4c86db9d15a489af7a12f31572f84fb745bae70bd48ed763e71fbebb3a7", "generation_seconds_batch": 22.894514483981766, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.256526470184326, "target_z": 1.2031250486359624, "ce": 5.617537498474121, "genuineness": 4.268857479095459, "identity": 0.47573643922805786, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuando estás lista.", "wer": 0.13333333333333333, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/f8c45be98be65ea1c70e545ecaa6030fd226ae49f79320ec08728f66dcba0355.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "fa69dc471f85ae1073d74e101187e5588d783d990becb83631748f97c8d574a0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2068438fe2172fd59f86c290267072a60b3e20b6bbb109f06222b81168687c7d", "generation_seconds_batch": 17.46393624099437, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.687375545501709, "target_z": 1.5156250702519456, "ce": 6.173524379730225, "genuineness": 4.118232727050781, "identity": 0.32872599363327026, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/fa69dc471f85ae1073d74e101187e5588d783d990becb83631748f97c8d574a0.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "fab83b07440a4b55eb30b687cd6315eb5cbcf503ea10fbe994c4030549fdca0c", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "7d9cf4fb1d4e86bde1f664c044ab88328d1e40cf26f7b95fd06e99bf3ea60dfa", "generation_seconds_batch": 18.49265929707326, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9028000831604004, "target_z": 1.6718750810599372, "ce": 5.681078910827637, "genuineness": 4.406930446624756, "identity": 0.5070608258247375, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás libre.", "wer": 0.06666666666666667, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/fab83b07440a4b55eb30b687cd6315eb5cbcf503ea10fbe994c4030549fdca0c.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "fb0ecf8eccd478706e9ef94f2805028a351b9b14df382d8b8e5378c6b760fd24", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "49ae4c04d2e9a1ffb66043f9244b8710de6ba265ea33cbbd06e28b95d5b62d60", "generation_seconds_batch": 24.09275859897025, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.405599594116211, "target_z": 0.5859375405299686, "ce": 6.332377910614014, "genuineness": 4.469690799713135, "identity": 0.6743133068084717, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/fb0ecf8eccd478706e9ef94f2805028a351b9b14df382d8b8e5378c6b760fd24.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "fb2fc2bead9185b326ab04ec8a04fb58c5c4df3ac6df434f1e40753f76f445ad", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1aab190dbe60b9b03428b37b6f3b3e9e172bac602b0d9d6e8486996fb72eb2fe", "generation_seconds_batch": 25.08338487497531, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8041348457336426, "target_z": 0.8749999567680335, "ce": 6.221180438995361, "genuineness": 4.770940780639648, "identity": 0.7005466222763062, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/fb2fc2bead9185b326ab04ec8a04fb58c5c4df3ac6df434f1e40753f76f445ad.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "fb7f593e32862126d5540492f00f8189dd437c1ec578fda3a0abb8c01aa90274", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "622b6327aa40789b7f3e59485fbf001bc450a98ed458ad3837bfc5850d9854b0", "generation_seconds_batch": 25.24895870196633, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.814906120300293, "target_z": 0.8828124918940062, "ce": 6.157639026641846, "genuineness": 4.281409740447998, "identity": 0.6363614201545715, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/fb7f593e32862126d5540492f00f8189dd437c1ec578fda3a0abb8c01aa90274.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "fb830326b0f998d7d348a1fc724e886bd99e45e271c744b19d39881d940a6f5b", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f001122030cb0bd51fea3803bb2e89c13f8bcb2c63c6752ba9bd7ca95c0f9ce8", "generation_seconds_batch": 23.1860186860431, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.085155487060547, "target_z": 0.35351557163554137, "ce": 6.205295085906982, "genuineness": 4.218649387359619, "identity": 0.7833038568496704, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/fb830326b0f998d7d348a1fc724e886bd99e45e271c744b19d39881d940a6f5b.mp3" }, { "id": "vn_S_ASMR_reg-de-0", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-0", "key": "fbf196a4b8ff63094a72c19900835b1169875ed5fd4724e2feeccf896691058d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Tür ist noch offen.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4ef5c1f39b755ca48715c0762a34b07d367d0e1d4c0a8c40f2cea773a67492c3", "generation_seconds_batch": 24.410459173028357, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3857812881469727, "target_z": 1.2968751242919037, "ce": 5.9431867599487305, "genuineness": 4.432034492492676, "identity": 0.44649356603622437, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/fbf196a4b8ff63094a72c19900835b1169875ed5fd4724e2feeccf896691058d.mp3" }, { "id": "vn_S_ASMR_reg-fr-0", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-0", "key": "fbf57df79ca44fe2c38a60576ba8c6e8e4ccd18d0fd456ee3e36b437aa029e5a", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La porte est encore ouverte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "490c9ec654df3db1c46a644a4251a0bdb56892f7bb31872711e070ba9a75052e", "generation_seconds_batch": 22.88479917193763, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8489439487457275, "target_z": 1.6328125783579392, "ce": 6.086154937744141, "genuineness": 4.381826400756836, "identity": 0.6332143545150757, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/fbf57df79ca44fe2c38a60576ba8c6e8e4ccd18d0fd456ee3e36b437aa029e5a.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "fc0d0d91f160fb310f1b7684f081e6e8f57aa4e320400bf35daa981f340189d5", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "362dc439fb46827ae4df347e0ab903a4338defcd78e10f76901b1b465eee335c", "generation_seconds_batch": 17.579466392984614, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6227481365203857, "target_z": 1.4687500324239748, "ce": 5.919358730316162, "genuineness": 4.432034492492676, "identity": 0.7045661211013794, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/fc0d0d91f160fb310f1b7684f081e6e8f57aa4e320400bf35daa981f340189d5.mp3" }, { "id": "vn_S_ASMR_reg-de-1", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-de-1", "key": "fce853395b440e8e9f982dceedb36926bc98c54f502b36349d5933cfa719da10", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d3bed8b6fe2f154696901630a7b8ab0a0a471bfd72420e9a983d1ad79149ca45", "generation_seconds_batch": 17.228487334912643, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2349839210510254, "target_z": 1.1874999783840168, "ce": 6.125868320465088, "genuineness": 4.419482707977295, "identity": 0.543371319770813, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/fce853395b440e8e9f982dceedb36926bc98c54f502b36349d5933cfa719da10.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "fd190efdac0034d93821b323e577d6fa0a49ad32e592b60d69999e71afa703af", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1c4ce8d8be2c29fdfe0e497a229c1731f75f7015edb74d4bc8d51aedef284eb3", "generation_seconds_batch": 24.330811747000553, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6317954063415527, "target_z": 0.7500000864639329, "ce": 6.062326908111572, "genuineness": 4.3881025314331055, "identity": 0.5654041767120361, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/fd190efdac0034d93821b323e577d6fa0a49ad32e592b60d69999e71afa703af.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "fd2f43375035f51d0d5e1f3b27d1daf29094f58983562d88435619321d3340dc", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "57a647a5ec914572939a039d9f6f57e9eb998d33c8fd81e63e0cacd52bef8261", "generation_seconds_batch": 18.013539887033403, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.364238739013672, "target_z": 1.281250054039958, "ce": 6.117925643920898, "genuineness": 4.469690799713135, "identity": 0.5819593667984009, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/fd2f43375035f51d0d5e1f3b27d1daf29094f58983562d88435619321d3340dc.mp3" }, { "id": "vn_S_ASMR_reg-nl-1", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-1", "key": "fda674acf94e5d9687779ead379acaf6cdb0251558acb9aa5472f3937beeea94", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Het stond naast het oude raam.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1e1f4fdc8fd3a8a40519df45c610cf43af7354b6341354531939fbb788d0890a", "generation_seconds_batch": 23.474032889003865, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.987245798110962, "target_z": 1.0078125351259728, "ce": 6.229123115539551, "genuineness": 4.532451152801514, "identity": 0.6397795677185059, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/fda674acf94e5d9687779ead379acaf6cdb0251558acb9aa5472f3937beeea94.mp3" }, { "id": "vn_S_ASMR_reg-fr-1", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-fr-1", "key": "fdae6822bd60bb95c3069a9f9b1a69e558bfe83e5f479650ccd52ad91436cd69", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "7584b0c71f3402773d84ebe7fe42a50e51505264893d722d9065f7d097e52248", "generation_seconds_batch": 24.63505137595348, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.838172674179077, "target_z": 1.6250000432319665, "ce": 5.903473377227783, "genuineness": 4.419482707977295, "identity": 0.8719573020935059, "hypothesis": "J'ai trouvé la petite boîte, elle était près de la vieille fenêtre, nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/fdae6822bd60bb95c3069a9f9b1a69e558bfe83e5f479650ccd52ad91436cd69.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "fdfb57ca3aa2512a6c6f5359242c10683b64fff7890c1678f1fdbb6e323a1567", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4cc02108d2b8315d526504891cb2e3dfec98025984e95457e6b505878bf15fa3", "generation_seconds_batch": 23.07523575297091, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1811277866363525, "target_z": 1.1484374756820188, "ce": 5.959072113037109, "genuineness": 4.444586753845215, "identity": 0.5307589769363403, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/fdfb57ca3aa2512a6c6f5359242c10683b64fff7890c1678f1fdbb6e323a1567.mp3" }, { "id": "vn_S_ASMR_reg-nl-0", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-nl-0", "key": "fe2d10969b19043898c3d448df2a10e5cfbf6280de9111204500baedb2f29b5e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] De deur staat nog open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "38cff6745751e6bf8b72f4c6f760edc635a98197734a281c68fddd801f088a59", "generation_seconds_batch": 24.37529632996302, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.202670097351074, "target_z": 1.1640623730060984, "ce": 5.689021587371826, "genuineness": 4.457139015197754, "identity": 0.35847264528274536, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/fe2d10969b19043898c3d448df2a10e5cfbf6280de9111204500baedb2f29b5e.mp3" }, { "id": "heldout-de-2", "language": "de", "sentences": [ "Die Straße führt hinter der Brücke nach links.", "Geh weiter... bis zu dem kleinen Holztor.", "Ich warte auf der anderen Seite." ], "text": "Die Straße führt hinter der Brücke nach links. Geh weiter... bis zu dem kleinen Holztor. Ich warte auf der anderen Seite.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-2", "key": "fe7fcd6329e62897e07fa2e41f121264f34b627646263c69da9e90c0e1134231", "bucket": "vn_S_ASMR_reg", "phase": "heldout", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Natural clear speech.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Natural clear speech.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 129, "seconds_generated": 10.32, "mp3_sha256": "72c96a8b81a52f47d43f534429605576d745e8b9b076bda1dcf54664b538678e", "generation_seconds_batch": 21.965058194007725, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4719510078430176, "target_z": 1.359375059443954, "ce": 5.744620323181152, "genuineness": 4.444586753845215, "identity": 0.46074485778808594, "hypothesis": "Die Straße führt hinter der Brücke nach links. Geh weiter bis zum kleinen Holztor. Ich warte auf der anderen Seite.", "wer": 0.09523809523809523, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/fe7fcd6329e62897e07fa2e41f121264f34b627646263c69da9e90c0e1134231.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "fee8cb6c456a376c353e396bcca5d42270ce20307fb1d055e921af43961ce64e", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b34e80f18c198a950aa5f0b7bcb7cac208ac3715540cc5c2c13247cbaf023531", "generation_seconds_batch": 23.83307847904507, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.31038236618042, "target_z": 1.2421873784100943, "ce": 6.197352409362793, "genuineness": 4.20609712600708, "identity": 0.5051141381263733, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/fee8cb6c456a376c353e396bcca5d42270ce20307fb1d055e921af43961ce64e.mp3" }, { "id": "vn_S_ASMR_reg-en-0", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-0", "key": "ff25a532bbd536940abd4d0ee9f2759b26a8bb65aea6cf421af1146906d94cc3", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] The door is still open.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] I put your coat on the chair.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b6498dba2aba23b1abfe4a6467a2e98b72854b852b4b6826b1106725abd79c16", "generation_seconds_batch": 24.908641059999354, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0949580669403076, "target_z": 1.0859375405299685, "ce": 6.133810997009277, "genuineness": 4.005263805389404, "identity": 0.45818835496902466, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ff25a532bbd536940abd4d0ee9f2759b26a8bb65aea6cf421af1146906d94cc3.mp3" }, { "id": "vn_S_ASMR_reg-en-1", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-en-1", "key": "ffbc945e16b395527edc5ce5f5780e186c9fc390ab9598e6dccbb8bf016c6220", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] I found the little box.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] It was beside the old window.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "945f47aaddc7d6ba3392ae81c89a440bde7307b8da9ca3246475a82a878569ae", "generation_seconds_batch": 15.354713826905936, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.901075839996338, "target_z": 0.9453124270460566, "ce": 6.229123115539551, "genuineness": 4.482243061065674, "identity": 0.516994833946228, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ffbc945e16b395527edc5ce5f5780e186c9fc390ab9598e6dccbb8bf016c6220.mp3" }, { "id": "vn_S_ASMR_reg-es-0", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-0", "key": "ffd496d6e451c005612c5a974741a72e8ad25b3a02c215630ef45b7b26e38d0d", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] La puerta sigue abierta.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "cfa5304d6cfa7ed15fd062bbf2d503afd1ff4050c084acb5ba6002e017d2b78c", "generation_seconds_batch": 23.535912253893912, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9532079696655273, "target_z": 0.2578124486620398, "ce": 6.0305562019348145, "genuineness": 4.080576419830322, "identity": 0.5155407786369324, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/ffd496d6e451c005612c5a974741a72e8ad25b3a02c215630ef45b7b26e38d0d.mp3" }, { "id": "vn_S_ASMR_reg-es-1", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "vn_S_ASMR_reg-es-1", "key": "fff81446ed53050c3fcbaad8df355b24e2625a2bebdd8e98252e436a48effdc0", "bucket": "vn_S_ASMR_reg", "phase": "grid", "candidate": "manual-763a50e7d58f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.\nSCRIPT:\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Establish it immediately. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] He encontrado la cajita.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Sustain the delivery. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Very close, gentle ASMR speech; breath detail and soft consonants, without mouth sound effects or added words. Preserve the intended style through the end. Begin with the feeling held back, then let it leak through and become unmistakable.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ea02eb37e6f44731a8241d36a05610020c7938ee4dbe121ed7ae87f10aecc3bc", "generation_seconds_batch": 22.32682138902601, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.204394340515137, "target_z": 1.8906250270199791, "ce": 5.959072113037109, "genuineness": 4.6705241203308105, "identity": 0.44998404383659363, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay donde.", "wer": 0.058823529411764705, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_ASMR_reg/fff81446ed53050c3fcbaad8df355b24e2625a2bebdd8e98252e436a48effdc0.mp3" }, { "key": "original-vn_S_ASMR_reg-de-0-seed1701-scale0.0", "bucket": "vn_S_ASMR_reg", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Tür ist noch offen.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "2873f5c23c5eea9d5ecc43787b358702de881c729f2cb5da3912c608c923544e", "task_id": "vn_S_ASMR_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Tür ist noch offen.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3527121543884277, "target_z": -0.17773439594048376, "ce": 6.284721851348877, "genuineness": 4.507347106933594, "identity": 0.5583904981613159, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/de/6709a1def273b7637517bd83.mp3" }, { "key": "original-vn_S_ASMR_reg-de-0-seed1701-scale0.5", "bucket": "vn_S_ASMR_reg", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Tür ist noch offen.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.5, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "59c5fec3742d55e15c21115b5fcc74a545fc4ca7251d6e9f885f8b1c6eada8f8", "task_id": "vn_S_ASMR_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Tür ist noch offen.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8185677528381348, "target_z": 0.160156278370978, "ce": 6.181467056274414, "genuineness": 4.607763767242432, "identity": 0.6890251636505127, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/de/4f44091813a47f8fe563ebe7.mp3" }, { "key": "original-vn_S_ASMR_reg-de-0-seed1701-scale0.75", "bucket": "vn_S_ASMR_reg", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Tür ist noch offen.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.75, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "c30a269a68a30c9abf2d00d9c56858b4d04bddedc1a7086f7b730ced509f7bd1", "task_id": "vn_S_ASMR_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Tür ist noch offen.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3352088928222656, "target_z": -0.19042972228822302, "ce": 6.284721851348877, "genuineness": 4.344170093536377, "identity": 0.3802715241909027, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/de/fa97056be99497dd5a2b18a0.mp3" }, { "key": "original-vn_S_ASMR_reg-de-0-seed1701-scale1.0", "bucket": "vn_S_ASMR_reg", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Tür ist noch offen.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 1.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "092ec6aca4ad08e116e2e062b2e84bdab8f1f336019e03fcb5f364b1d008984a", "task_id": "vn_S_ASMR_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Tür ist noch offen.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2907776832580566, "target_z": -0.22265621352302828, "ce": 6.284721851348877, "genuineness": 4.293961524963379, "identity": 0.5212423205375671, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/de/faeea053ae6d023007490920.mp3" }, { "key": "original-vn_S_ASMR_reg-de-0-seed1702-scale0.0", "bucket": "vn_S_ASMR_reg", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Tür ist noch offen.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "828655557a18ee4742094acec3cec899e97fe51fcd047ba7cdc3bdf1a6171b7b", "task_id": "vn_S_ASMR_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Tür ist noch offen.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4756126403808594, "target_z": 0.6367188459209256, "ce": 6.292664527893066, "genuineness": 4.494794845581055, "identity": 0.5078657865524292, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/de/2516cdb5c8806e689ec4b6fb.mp3" }, { "key": "original-vn_S_ASMR_reg-de-0-seed1702-scale0.5", "bucket": "vn_S_ASMR_reg", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Tür ist noch offen.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.5, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "95434b705bdd5f6f78770ba3a41cd6a9321e629b44a441568bc611be977f0e9c", "task_id": "vn_S_ASMR_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Tür ist noch offen.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3823330402374268, "target_z": -0.1562500108079916, "ce": 6.252951145172119, "genuineness": 4.293961524963379, "identity": 0.4750843942165375, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/de/9592a29113e9febc63bca731.mp3" }, { "key": "original-vn_S_ASMR_reg-de-0-seed1702-scale0.75", "bucket": "vn_S_ASMR_reg", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Tür ist noch offen.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.75, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "3ff60118b6ae313f920944d53d748ca26b9119b8ea237b3c60eb01ebbfef826d", "task_id": "vn_S_ASMR_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Tür ist noch offen.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.76104998588562, "target_z": 0.8437499891920084, "ce": 6.205295085906982, "genuineness": 4.532451152801514, "identity": 0.7036466598510742, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/de/054922269466aa24d64014f6.mp3" }, { "key": "original-vn_S_ASMR_reg-de-0-seed1702-scale1.0", "bucket": "vn_S_ASMR_reg", "language": "de", "sentences": [ "Die Tür ist noch offen.", "Ich habe deinen Mantel auf den Stuhl gelegt.", "Sag mir, wenn du bereit bist." ], "text": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Tür ist noch offen.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 1.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "b76d5ff39d283ca9be55245fcd4e36b58f30bbd0b388853c8d59951445962eda", "task_id": "vn_S_ASMR_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Die Tür ist noch offen.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9020447731018066, "target_z": 0.22070316620546804, "ce": 6.205295085906982, "genuineness": 4.444586753845215, "identity": 0.6176893711090088, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/de/99fa28d45afc730800ebd92b.mp3" }, { "key": "original-vn_S_ASMR_reg-de-1-seed1701-scale0.0", "bucket": "vn_S_ASMR_reg", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "893dee8e79fefc209ebd17ed5021b579e3651642f5d9d67a65651c2019e6dd16", "task_id": "vn_S_ASMR_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2800064086914062, "target_z": -0.23046874864900105, "ce": 6.316492557525635, "genuineness": 4.2751336097717285, "identity": 0.4087516665458679, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/de/a4046474362f564c74f28e92.mp3" }, { "key": "original-vn_S_ASMR_reg-de-1-seed1701-scale0.5", "bucket": "vn_S_ASMR_reg", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.5, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "09d207df889fe29e937dd8aca654eeff905ee9ba189bbd4dbffbccdf2a55b768", "task_id": "vn_S_ASMR_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8024108409881592, "target_z": 0.14843747568201884, "ce": 6.2767791748046875, "genuineness": 4.394378662109375, "identity": 0.3573176860809326, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/de/a428acf6321f3132ab88064d.mp3" }, { "key": "original-vn_S_ASMR_reg-de-1-seed1701-scale0.75", "bucket": "vn_S_ASMR_reg", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.75, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "4ecc0e0549c041180967af40d0d0cff73de858f1ea4e3f8b5f0f2e69135862e8", "task_id": "vn_S_ASMR_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7997180223464966, "target_z": 0.14648434190052567, "ce": 6.3641486167907715, "genuineness": 4.256305694580078, "identity": 0.572624921798706, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/de/40f16000318b4571e9e70044.mp3" }, { "key": "original-vn_S_ASMR_reg-de-1-seed1701-scale1.0", "bucket": "vn_S_ASMR_reg", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 1.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "e3006679dae556ed09d820ed0bf168779fa2471538bf67b768f87b850108100f", "task_id": "vn_S_ASMR_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2207646369934082, "target_z": -0.2734375189139853, "ce": 6.395919322967529, "genuineness": 4.2751336097717285, "identity": 0.48834148049354553, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/de/581765e9a21fbb0bca4ef22c.mp3" }, { "key": "original-vn_S_ASMR_reg-de-1-seed1702-scale0.0", "bucket": "vn_S_ASMR_reg", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "dd6a0f285f956f05e675f664848ac9a2919f2473efdd6a80551806965385624c", "task_id": "vn_S_ASMR_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2073006629943848, "target_z": -0.2832031013575183, "ce": 6.427689552307129, "genuineness": 4.3504462242126465, "identity": 0.5189880728721619, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/de/7c2886a2b14fd7bf5b6ec076.mp3" }, { "key": "original-vn_S_ASMR_reg-de-1-seed1702-scale0.5", "bucket": "vn_S_ASMR_reg", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.5, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "9d31499f2e5e3aa171271d02e0dd5b416c04480adf644651f28be604cd7981ea", "task_id": "vn_S_ASMR_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.274620771408081, "target_z": -0.23437501621198742, "ce": 6.459460258483887, "genuineness": 4.519899368286133, "identity": 0.5974308252334595, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/de/2078764e13b9cb3763742ace.mp3" }, { "key": "original-vn_S_ASMR_reg-de-1-seed1702-scale0.75", "bucket": "vn_S_ASMR_reg", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.75, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "650c68cf8b15bf3733292ff8874a6a351f273c02f5834f0c20600f91e3408d91", "task_id": "vn_S_ASMR_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8643454313278198, "target_z": 0.19335937972849634, "ce": 6.348263263702393, "genuineness": 4.406930446624756, "identity": 0.5251948237419128, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/de/52166572ddbe649ebd4a7d89.mp3" }, { "key": "original-vn_S_ASMR_reg-de-1-seed1702-scale1.0", "bucket": "vn_S_ASMR_reg", "language": "de", "sentences": [ "Ich habe die kleine Schachtel gefunden.", "Sie stand neben dem alten Fenster.", "Wir können gemeinsam hineinschauen." ], "text": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 1.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "43954966456249775924638aeb277e3b997afaa72aa415a2e78a05e257ef6754", "task_id": "vn_S_ASMR_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1847894191741943, "target_z": 0.4257812621589906, "ce": 6.348263263702393, "genuineness": 4.557555198669434, "identity": 0.6454848051071167, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/de/1e6277762be897f1e0fd86a7.mp3" }, { "key": "original-vn_S_ASMR_reg-en-0-seed1701-scale0.0", "bucket": "vn_S_ASMR_reg", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The door is still open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I put your coat on the chair.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.0, "frames_generated": 127, "seconds_generated": 10.16, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "7c37d17fcf48dca6fe68326c3bc21ccf82f3a33a6a7a221728e44974e5ec2fea", "task_id": "vn_S_ASMR_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The door is still open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I put your coat on the chair.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8333781957626343, "target_z": 0.17089847093722407, "ce": 6.173524379730225, "genuineness": 3.8295347690582275, "identity": 0.24234101176261902, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/en/8e058b9cce28380de36c6668.mp3" }, { "key": "original-vn_S_ASMR_reg-en-0-seed1701-scale0.5", "bucket": "vn_S_ASMR_reg", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The door is still open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I put your coat on the chair.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.5, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "be6c1fc1babf85d62f7748ac209ce2bd716f969536cdd51f4cd09c8fa2f7d793", "task_id": "vn_S_ASMR_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The door is still open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I put your coat on the chair.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6802659034729004, "target_z": 0.7851563216029445, "ce": 6.284721851348877, "genuineness": 4.193545341491699, "identity": 0.13690945506095886, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/en/520c4bc8e66e7a01a9a9395f.mp3" }, { "key": "original-vn_S_ASMR_reg-en-0-seed1701-scale0.75", "bucket": "vn_S_ASMR_reg", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The door is still open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I put your coat on the chair.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.75, "frames_generated": 127, "seconds_generated": 10.16, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "6169452935aab2d3462398d1eda14322222c3d0350c59e8d26580b25e550121a", "task_id": "vn_S_ASMR_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The door is still open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I put your coat on the chair.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4271419048309326, "target_z": 0.6015624378540482, "ce": 6.268836498260498, "genuineness": 4.344170093536377, "identity": 0.41686421632766724, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/en/d7a58778e27490ae9f3d38fb.mp3" }, { "key": "original-vn_S_ASMR_reg-en-0-seed1701-scale1.0", "bucket": "vn_S_ASMR_reg", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The door is still open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I put your coat on the chair.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 1.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "565c7994073079719f02f04781cee64e754160d5c4a8fe61a0b10aee9dae7f24", "task_id": "vn_S_ASMR_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The door is still open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I put your coat on the chair.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8226069211959839, "target_z": 0.1630859358112513, "ce": 6.24500846862793, "genuineness": 4.074300289154053, "identity": 0.5661787986755371, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/en/d844511097b306864ce5e88e.mp3" }, { "key": "original-vn_S_ASMR_reg-en-0-seed1702-scale0.0", "bucket": "vn_S_ASMR_reg", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The door is still open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I put your coat on the chair.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "ad54cb9471e3a9c4ef97f5bd4d80c44b366652e5f46b3c24d10079fc92672da9", "task_id": "vn_S_ASMR_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The door is still open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I put your coat on the chair.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4267643690109253, "target_z": -0.1240234331092534, "ce": 6.149696350097656, "genuineness": 3.8295347690582275, "identity": 0.26559942960739136, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/en/5a5141b1f5014fe7eb6911c8.mp3" }, { "key": "original-vn_S_ASMR_reg-en-0-seed1702-scale0.5", "bucket": "vn_S_ASMR_reg", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The door is still open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I put your coat on the chair.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.5, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "e93835b501ee12a35ab19002fc37e9f04ddbe9535cc4f54ce55e955780da7649", "task_id": "vn_S_ASMR_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The door is still open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I put your coat on the chair.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.346357822418213, "target_z": 0.5429687702649842, "ce": 6.109982967376709, "genuineness": 4.13078498840332, "identity": 0.38558146357536316, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/en/5f4d3162ab18be29cdf9e402.mp3" }, { "key": "original-vn_S_ASMR_reg-en-0-seed1702-scale0.75", "bucket": "vn_S_ASMR_reg", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The door is still open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I put your coat on the chair.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.75, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "6b9e4a5f0de3d8a6328e41364e6ab29951ce4edd61dede1cdec09c2ffaeeff40", "task_id": "vn_S_ASMR_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The door is still open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I put your coat on the chair.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9828288555145264, "target_z": 0.27929683379453196, "ce": 6.213237762451172, "genuineness": 3.7793264389038086, "identity": 0.16851991415023804, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/en/c84ee3dc4134e4eda29082a4.mp3" }, { "key": "original-vn_S_ASMR_reg-en-0-seed1702-scale1.0", "bucket": "vn_S_ASMR_reg", "language": "en", "sentences": [ "The door is still open.", "I put your coat on the chair.", "Tell me when you are ready." ], "text": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The door is still open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I put your coat on the chair.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 1.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "f19ebb7849912f387241464bad5297a3f364c4ad508126e82e433afffda911c0", "task_id": "vn_S_ASMR_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] The door is still open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] I put your coat on the chair.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3625144958496094, "target_z": 0.5546874000260775, "ce": 6.205295085906982, "genuineness": 4.036643981933594, "identity": 0.23480314016342163, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/en/a8a930ab8ca2f7ac01477097.mp3" }, { "key": "original-vn_S_ASMR_reg-en-1-seed1701-scale0.0", "bucket": "vn_S_ASMR_reg", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I found the little box.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It was beside the old window.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can look inside together.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.0, "frames_generated": 127, "seconds_generated": 10.16, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "728432ec4697a48067ad9504bdff03d94d285e3313ada1ea551c6a6339334d6c", "task_id": "vn_S_ASMR_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I found the little box.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It was beside the old window.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can look inside together.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6489208936691284, "target_z": 0.03710936892050471, "ce": 6.403861999511719, "genuineness": 4.3881025314331055, "identity": 0.37112343311309814, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/en/4eefed4ff48eefb6f456a1b2.mp3" }, { "key": "original-vn_S_ASMR_reg-en-1-seed1701-scale0.5", "bucket": "vn_S_ASMR_reg", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I found the little box.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It was beside the old window.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can look inside together.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.5, "frames_generated": 127, "seconds_generated": 10.16, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "58a102afbc97ea4c2a6949f0b7f0c5bdaf7f614173e46010f9930bb473632399", "task_id": "vn_S_ASMR_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I found the little box.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It was beside the old window.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can look inside together.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7889468669891357, "target_z": 0.13867189323848586, "ce": 6.324435234069824, "genuineness": 4.582659721374512, "identity": 0.4757991135120392, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/en/18a0554ddb295c14eaffea3b.mp3" }, { "key": "original-vn_S_ASMR_reg-en-1-seed1701-scale0.75", "bucket": "vn_S_ASMR_reg", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I found the little box.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It was beside the old window.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can look inside together.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.75, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "cacd873f4fc8388f8c6d05015c6fe742663b263aa5e982e9c32731267741cbfa", "task_id": "vn_S_ASMR_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I found the little box.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It was beside the old window.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can look inside together.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1901750564575195, "target_z": 0.42968752972197694, "ce": 6.443574905395508, "genuineness": 4.6203155517578125, "identity": 0.4900534152984619, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/en/830acea8268d44d67f13a54c.mp3" }, { "key": "original-vn_S_ASMR_reg-en-1-seed1701-scale1.0", "bucket": "vn_S_ASMR_reg", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I found the little box.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It was beside the old window.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can look inside together.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 1.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "eaae9032284e08c65f213377598b059d6c316f68c83b1c095b4d6e2e266ae871", "task_id": "vn_S_ASMR_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I found the little box.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It was beside the old window.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can look inside together.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.147090196609497, "target_z": 0.39843756214595183, "ce": 6.435632705688477, "genuineness": 4.846253395080566, "identity": 0.4225428104400635, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/en/bdc52096280df00babd3c10d.mp3" }, { "key": "original-vn_S_ASMR_reg-en-1-seed1702-scale0.0", "bucket": "vn_S_ASMR_reg", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I found the little box.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It was beside the old window.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can look inside together.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "251745320f22b354b00f2b67eb837f9c457dbd2085e3be2ed8a84a080c3e1da0", "task_id": "vn_S_ASMR_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I found the little box.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It was beside the old window.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can look inside together.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3984899520874023, "target_z": -0.14453120811903247, "ce": 6.252951145172119, "genuineness": 4.545003414154053, "identity": 0.4173415005207062, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/en/8dc902cf6ed30e3da2cadb53.mp3" }, { "key": "original-vn_S_ASMR_reg-en-1-seed1702-scale0.5", "bucket": "vn_S_ASMR_reg", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I found the little box.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It was beside the old window.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can look inside together.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.5, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "742ce53f3956198eae19534f9f1b42c0d9929ef3cd3b81a62b0dc95e10d08540", "task_id": "vn_S_ASMR_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I found the little box.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It was beside the old window.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can look inside together.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5492870807647705, "target_z": -0.035156235139011525, "ce": 6.332377910614014, "genuineness": 4.532451152801514, "identity": 0.3800475597381592, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/en/f00422e8713577a0240d1e81.mp3" }, { "key": "original-vn_S_ASMR_reg-en-1-seed1702-scale0.75", "bucket": "vn_S_ASMR_reg", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I found the little box.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It was beside the old window.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can look inside together.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.75, "frames_generated": 127, "seconds_generated": 10.16, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "862f6dd903470cd0946fc659fbd615a66707d2a8b0ee07c64b313635df7b3a82", "task_id": "vn_S_ASMR_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I found the little box.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It was beside the old window.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can look inside together.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9612864255905151, "target_z": 0.2636718500065194, "ce": 6.427689552307129, "genuineness": 4.419482707977295, "identity": 0.35325610637664795, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/en/31ed58ae9460de4ee6f191fe.mp3" }, { "key": "original-vn_S_ASMR_reg-en-1-seed1702-scale1.0", "bucket": "vn_S_ASMR_reg", "language": "en", "sentences": [ "I found the little box.", "It was beside the old window.", "We can look inside together." ], "text": "I found the little box. It was beside the old window. We can look inside together.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I found the little box.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It was beside the old window.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can look inside together.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 1.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "99bb1833069a74c890ad7a51bc0390185f36996a5a0c0bae2ed3c756f8ac07bd", "task_id": "vn_S_ASMR_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] I found the little box.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] It was beside the old window.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can look inside together.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8212605714797974, "target_z": 0.16210941215247118, "ce": 6.38003396987915, "genuineness": 4.6203155517578125, "identity": 0.4257907271385193, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/en/c9fb1d58346a526c80196c4a.mp3" }, { "key": "original-vn_S_ASMR_reg-es-0-seed1701-scale0.0", "bucket": "vn_S_ASMR_reg", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "2b4dfd90fb771b1e9567bd982d700dae9c243d5d3639a9867bcf8b4f0edfb6ae", "task_id": "vn_S_ASMR_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4540700912475586, "target_z": 0.62109377566898, "ce": 5.911416053771973, "genuineness": 4.545003414154053, "identity": 0.6398885250091553, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/es/f6dc79753f0a6b1855613ef5.mp3" }, { "key": "original-vn_S_ASMR_reg-es-0-seed1701-scale0.5", "bucket": "vn_S_ASMR_reg", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.5, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "540f90b0120a97d93b4f435a068f69995a1ad4aa09c0d29422021c2ee8541cb0", "task_id": "vn_S_ASMR_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3032729625701904, "target_z": 0.5117188026889592, "ce": 6.046441555023193, "genuineness": 4.356721878051758, "identity": 0.456123948097229, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/es/45d86a689185450d32462cf4.mp3" }, { "key": "original-vn_S_ASMR_reg-es-0-seed1701-scale0.75", "bucket": "vn_S_ASMR_reg", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.75, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "97bb9ce87023936b12c15204d2f4c13f47289a48d4a547f7a716ff53e3ff6932", "task_id": "vn_S_ASMR_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2611567974090576, "target_z": -0.24414059865552043, "ce": 5.990842819213867, "genuineness": 3.948779582977295, "identity": 0.5969647169113159, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/es/7ad1490246631799e7708753.mp3" }, { "key": "original-vn_S_ASMR_reg-es-0-seed1701-scale1.0", "bucket": "vn_S_ASMR_reg", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 1.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "54b961b8299243e1790cc4efcd5359ac9ff9f227cec181ec8657174ff131cf6a", "task_id": "vn_S_ASMR_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3086585998535156, "target_z": 0.5156250702519455, "ce": 6.1894097328186035, "genuineness": 4.344170093536377, "identity": 0.5455687046051025, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/es/24829a2bb294fff37cd5d005.mp3" }, { "key": "original-vn_S_ASMR_reg-es-0-seed1702-scale0.0", "bucket": "vn_S_ASMR_reg", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "dbc7223f48c5a9d8c2c3baf736aaf1052091ca0f9e64bc652cc6b72fb8fe48ad", "task_id": "vn_S_ASMR_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9909073114395142, "target_z": 0.2851562351390115, "ce": 5.990842819213867, "genuineness": 4.093128204345703, "identity": 0.4915448725223541, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/es/909ce01bc5f67bb5de7b50d1.mp3" }, { "key": "original-vn_S_ASMR_reg-es-0-seed1702-scale0.5", "bucket": "vn_S_ASMR_reg", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.5, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "8a1d4f5edd4fcf6705779291630519524520880d4cbf98b5daa82f63cc0bf425", "task_id": "vn_S_ASMR_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.621992826461792, "target_z": 0.017578117569505763, "ce": 5.935244083404541, "genuineness": 4.074300289154053, "identity": 0.6436977386474609, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/es/da69a4dc86147efaa4770f97.mp3" }, { "key": "original-vn_S_ASMR_reg-es-0-seed1702-scale0.75", "bucket": "vn_S_ASMR_reg", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.75, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "3fcb6051950df13a91885498269b8cf4a8e3c990515828ecdd08883a2d84072d", "task_id": "vn_S_ASMR_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.76104998588562, "target_z": 0.8437499891920084, "ce": 6.0146708488464355, "genuineness": 4.143336772918701, "identity": 0.5188066959381104, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/es/dff54ff41de4942f9cc8cf8d.mp3" }, { "key": "original-vn_S_ASMR_reg-es-0-seed1702-scale1.0", "bucket": "vn_S_ASMR_reg", "language": "es", "sentences": [ "La puerta sigue abierta.", "He dejado tu abrigo sobre la silla.", "Dime cuándo estás listo." ], "text": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 1.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "a94593dc97096ace838e2dec463424fe9e7c2c3b4b8fb34a150d93d4a4d10510", "task_id": "vn_S_ASMR_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La puerta sigue abierta.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8037573099136353, "target_z": 0.14941408580473192, "ce": 5.998785495758057, "genuineness": 3.998987913131714, "identity": 0.318739652633667, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/es/462192b4ffda256d82eaf3af.mp3" }, { "key": "original-vn_S_ASMR_reg-es-1-seed1701-scale0.0", "bucket": "vn_S_ASMR_reg", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "75e909e856a91e139854ae436440d5e7c8da55c455a87e72ec197f032e16ca92", "task_id": "vn_S_ASMR_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.358097791671753, "target_z": -0.1738281283774974, "ce": 5.95112943649292, "genuineness": 4.162164688110352, "identity": 0.7271223068237305, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/es/b4f6cdac372c9e57d7b907bb.mp3" }, { "key": "original-vn_S_ASMR_reg-es-1-seed1701-scale0.5", "bucket": "vn_S_ASMR_reg", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.5, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "02ed80098574af8ea5602281a7b076dfb8ee15ac485c753fd97e75ca81ab039e", "task_id": "vn_S_ASMR_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.928004026412964, "target_z": 0.9648437648609884, "ce": 6.292664527893066, "genuineness": 4.6705241203308105, "identity": 0.5831764936447144, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/es/a349b24bc2bcca07b1bb533b.mp3" }, { "key": "original-vn_S_ASMR_reg-es-1-seed1701-scale0.75", "bucket": "vn_S_ASMR_reg", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.75, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "d4c82f1df2d162004d3d4f674639f17ae2f54c181fd5d9f1b4a2854229b50e36", "task_id": "vn_S_ASMR_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6048672199249268, "target_z": 0.730468748649001, "ce": 6.205295085906982, "genuineness": 4.65797233581543, "identity": 0.5398087501525879, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/es/293c32a153b367d188ddaa8c.mp3" }, { "key": "original-vn_S_ASMR_reg-es-1-seed1701-scale1.0", "bucket": "vn_S_ASMR_reg", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 1.0, "frames_generated": 127, "seconds_generated": 10.16, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "a285b42b208164991ade77562bb6c3c48dfde529a0ccd17a980ea0d1a1614ff5", "task_id": "vn_S_ASMR_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0411019325256348, "target_z": 1.0468750378279708, "ce": 6.1020402908325195, "genuineness": 4.758388519287109, "identity": 0.7223674058914185, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/es/5b21aca2c2f1feff9b3e82dd.mp3" }, { "key": "original-vn_S_ASMR_reg-es-1-seed1702-scale0.0", "bucket": "vn_S_ASMR_reg", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "311e2bb4bad858cc3ddc1d17fb0c1e130b98135e61c886ef7dc577dc6e34f425", "task_id": "vn_S_ASMR_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.865691900253296, "target_z": 0.1943359898512094, "ce": 6.125868320465088, "genuineness": 4.457139015197754, "identity": 0.3941265344619751, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/es/1d9c1481803921fdb4a405f0.mp3" }, { "key": "original-vn_S_ASMR_reg-es-1-seed1702-scale0.5", "bucket": "vn_S_ASMR_reg", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.5, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "e7eac442f30d6e0878d694272813d63d43284aa8b54630b3b03b80296c22f20b", "task_id": "vn_S_ASMR_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0178353786468506, "target_z": 0.3046874864900105, "ce": 6.229123115539551, "genuineness": 4.337893962860107, "identity": 0.6367200016975403, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/es/3a88b742805da6d1543445bd.mp3" }, { "key": "original-vn_S_ASMR_reg-es-1-seed1702-scale0.75", "bucket": "vn_S_ASMR_reg", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.75, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "b469d7dd7719a1be5b819db251cf4f9cc7bd66fe4e4c84b3937d100ea64885f1", "task_id": "vn_S_ASMR_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1847894191741943, "target_z": 0.4257812621589906, "ce": 6.165581703186035, "genuineness": 4.432034492492676, "identity": 0.5253204703330994, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/es/3eff3ab645ab82b30abb7436.mp3" }, { "key": "original-vn_S_ASMR_reg-es-1-seed1702-scale1.0", "bucket": "vn_S_ASMR_reg", "language": "es", "sentences": [ "He encontrado la cajita.", "Estaba junto a la ventana vieja.", "Podemos mirar juntos lo que hay dentro." ], "text": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 1.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "d43c8f693b0a329bfb06ecbedb611e320e1acab30b37268a068f2c21cbc0ba14", "task_id": "vn_S_ASMR_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] He encontrado la cajita.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5735223293304443, "target_z": -0.017578117569505763, "ce": 5.863759994506836, "genuineness": 4.400654315948486, "identity": 0.7073044776916504, "hypothesis": "He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/es/2339e74c7243979a24558cae.mp3" }, { "key": "original-vn_S_ASMR_reg-fr-0-seed1701-scale0.0", "bucket": "vn_S_ASMR_reg", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La porte est encore ouverte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "c1f8de89456190d93309c1c3cd6367443b3e9b96961766572faa54a0be6423a3", "task_id": "vn_S_ASMR_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La porte est encore ouverte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4540700912475586, "target_z": 0.62109377566898, "ce": 6.09409761428833, "genuineness": 4.758388519287109, "identity": 0.7359107732772827, "hypothesis": "La porte est encore ouverte, j'ai posé ton manteau sur la chaise, dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/fr/7c30d6f80a9cf6ba366bf6c1.mp3" }, { "key": "original-vn_S_ASMR_reg-fr-0-seed1701-scale0.5", "bucket": "vn_S_ASMR_reg", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La porte est encore ouverte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.5, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "e90d3638d53f8485dcaa1102394de09fd6e4d2d3c48832cd1ba5b96eb2ceac48", "task_id": "vn_S_ASMR_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La porte est encore ouverte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7125794887542725, "target_z": 0.8085937540529968, "ce": 6.149696350097656, "genuineness": 4.607763767242432, "identity": 0.6657980680465698, "hypothesis": "La porte est encore ouverte, j'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/fr/3b27673d3b97695bdd790e52.mp3" }, { "key": "original-vn_S_ASMR_reg-fr-0-seed1701-scale0.75", "bucket": "vn_S_ASMR_reg", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La porte est encore ouverte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.75, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "689814a3c506a611ef18bf2831b8c7cc12138162d51a5cccee9546282f4286a4", "task_id": "vn_S_ASMR_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La porte est encore ouverte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5581207275390625, "target_z": 1.421874994596004, "ce": 6.173524379730225, "genuineness": 4.532451152801514, "identity": 0.6764954328536987, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/fr/fd7274ae8058ffcfd473673d.mp3" }, { "key": "original-vn_S_ASMR_reg-fr-0-seed1701-scale1.0", "bucket": "vn_S_ASMR_reg", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La porte est encore ouverte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 1.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "79f24da5d22a32d9175e860b71dcef1a833be212f7e468008becc63745ec1a1a", "task_id": "vn_S_ASMR_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La porte est encore ouverte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.227874279022217, "target_z": 0.4570312297350157, "ce": 6.23706579208374, "genuineness": 4.532451152801514, "identity": 0.7452595829963684, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/fr/b86aefd3a1a95cdbda7a5b69.mp3" }, { "key": "original-vn_S_ASMR_reg-fr-0-seed1702-scale0.0", "bucket": "vn_S_ASMR_reg", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La porte est encore ouverte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "82e7f2f3c1afd3fa60521b4199d562811d48ba0dbab635230bf35d3b10ce4686", "task_id": "vn_S_ASMR_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La porte est encore ouverte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9235872030258179, "target_z": 0.23632814999348062, "ce": 6.078212261199951, "genuineness": 3.9550557136535645, "identity": 0.6503573060035706, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/fr/0b64671e50a166007795aa13.mp3" }, { "key": "original-vn_S_ASMR_reg-fr-0-seed1702-scale0.5", "bucket": "vn_S_ASMR_reg", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La porte est encore ouverte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.5, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "c1f4195d6e314ca6569260778db1d22650b3da175ce603a77c271c70cf63879b", "task_id": "vn_S_ASMR_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La porte est encore ouverte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.443298816680908, "target_z": 0.6132812405430074, "ce": 6.252951145172119, "genuineness": 4.319066047668457, "identity": 0.7366520166397095, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/fr/4fe46f294af81d572f74b087.mp3" }, { "key": "original-vn_S_ASMR_reg-fr-0-seed1702-scale0.75", "bucket": "vn_S_ASMR_reg", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La porte est encore ouverte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.75, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "eda4de724dc50a63d8fee64738da81479b56d85808ad0f43f6083760a7da9e65", "task_id": "vn_S_ASMR_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La porte est encore ouverte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.518697500228882, "target_z": 0.6679688134969508, "ce": 6.181467056274414, "genuineness": 4.6203155517578125, "identity": 0.6983445286750793, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/fr/ed02f9ca94a064634c4e2c02.mp3" }, { "key": "original-vn_S_ASMR_reg-fr-0-seed1702-scale1.0", "bucket": "vn_S_ASMR_reg", "language": "fr", "sentences": [ "La porte est encore ouverte.", "J’ai posé ton manteau sur la chaise.", "Dis-moi quand tu es prêt." ], "text": "La porte est encore ouverte. J’ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La porte est encore ouverte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 1.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "de8700e5bfe9f5b7bda31c62fa439ea97451776dadd4a9aec0fbd09daa246af2", "task_id": "vn_S_ASMR_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] La porte est encore ouverte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2349839210510254, "target_z": 1.1874999783840168, "ce": 6.046441555023193, "genuineness": 4.432034492492676, "identity": 0.8543679118156433, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/fr/be55dabbea949ce33c1e7f5d.mp3" }, { "key": "original-vn_S_ASMR_reg-fr-1-seed1701-scale0.0", "bucket": "vn_S_ASMR_reg", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "4f2d3a8356e8a7e5b6ac556ab2a4e561891e4e3a935cfced2e4f03055ffa6aa2", "task_id": "vn_S_ASMR_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1093907356262207, "target_z": 0.37109368920504715, "ce": 6.24500846862793, "genuineness": 4.808597087860107, "identity": 0.8227828145027161, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/fr/57bd6c7e9b3997287eeeff0f.mp3" }, { "key": "original-vn_S_ASMR_reg-fr-1-seed1701-scale0.5", "bucket": "vn_S_ASMR_reg", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.5, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "0f30f5064fcea468c0f74a85b15c7c8ee628c2b6d3f2e73efb74aeade86b6a89", "task_id": "vn_S_ASMR_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6317954063415527, "target_z": 0.7500000864639329, "ce": 6.181467056274414, "genuineness": 4.65797233581543, "identity": 0.779592752456665, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/fr/3215c3fa70f17e608f290772.mp3" }, { "key": "original-vn_S_ASMR_reg-fr-1-seed1701-scale0.75", "bucket": "vn_S_ASMR_reg", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.75, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "ca6cc8b2225321853f20bccaeaaeb9df01c86984b833f0e673c470dfa2e35f90", "task_id": "vn_S_ASMR_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.019559383392334, "target_z": 1.0312499675760252, "ce": 6.1894097328186035, "genuineness": 4.532451152801514, "identity": 0.7140898704528809, "hypothesis": "J'ai trouvé la petite boîte, elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/fr/67b53964ab5d5aea2dc6f14d.mp3" }, { "key": "original-vn_S_ASMR_reg-fr-1-seed1701-scale1.0", "bucket": "vn_S_ASMR_reg", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 1.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "f52b32205fde8a31b06c11ba243bb98601e81c95434415414b90d47d856bf0a4", "task_id": "vn_S_ASMR_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5994815826416016, "target_z": 0.7265624810860146, "ce": 6.252951145172119, "genuineness": 4.65797233581543, "identity": 0.7965625524520874, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/fr/fdd381fe0faa8516195d0fc2.mp3" }, { "key": "original-vn_S_ASMR_reg-fr-1-seed1702-scale0.0", "bucket": "vn_S_ASMR_reg", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "fb9b984905d0b049e27e1dc70f12b3443018720e4136cbe4af810db71a884581", "task_id": "vn_S_ASMR_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.085155487060547, "target_z": 0.35351557163554137, "ce": 6.260893821716309, "genuineness": 4.532451152801514, "identity": 0.7282161712646484, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/fr/5e545dd8cc9e844ae9d51100.mp3" }, { "key": "original-vn_S_ASMR_reg-fr-1-seed1702-scale0.5", "bucket": "vn_S_ASMR_reg", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.5, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "7f536a98abb323d0f1c78707c7b21fdca8738ab3932bea4ec5cc22885846b48e", "task_id": "vn_S_ASMR_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2413382530212402, "target_z": 0.46679681217854874, "ce": 6.252951145172119, "genuineness": 4.419482707977295, "identity": 0.6927187442779541, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/fr/d8f5e5b6679e834e011a750f.mp3" }, { "key": "original-vn_S_ASMR_reg-fr-1-seed1702-scale0.75", "bucket": "vn_S_ASMR_reg", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.75, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "68d3290274e30f20445def046097d2ae762a5e75448917ef62b7da48106df4c0", "task_id": "vn_S_ASMR_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.459455728530884, "target_z": 0.6250000432319665, "ce": 6.157639026641846, "genuineness": 4.331617832183838, "identity": 0.7466367483139038, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/fr/d32a1b6ae1b9156e6f2bd4d1.mp3" }, { "key": "original-vn_S_ASMR_reg-fr-1-seed1702-scale1.0", "bucket": "vn_S_ASMR_reg", "language": "fr", "sentences": [ "J’ai trouvé la petite boîte.", "Elle était près de la vieille fenêtre.", "Nous pouvons regarder ensemble ce qu’elle contient." ], "text": "J’ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu’elle contient.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 1.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "916c4835251a5e1d5a66998db32e74cc6d19c99784c3b4ed7091aa05748785d1", "task_id": "vn_S_ASMR_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2197959423065186, "target_z": 0.4511719148544691, "ce": 6.070269584655762, "genuineness": 4.457139015197754, "identity": 0.6142390966415405, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/fr/2d8c8c35bb8c77f9de559105.mp3" }, { "key": "original-vn_S_ASMR_reg-nl-0-seed1701-scale0.0", "bucket": "vn_S_ASMR_reg", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De deur staat nog open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "14e33983615f757e4896497e70d27743443d8ad673f5d1d99645f302ccbee4d7", "task_id": "vn_S_ASMR_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De deur staat nog open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0645818710327148, "target_z": -0.3867187594569927, "ce": 6.046441555023193, "genuineness": 4.400654315948486, "identity": 0.6846068501472473, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/nl/aea25e77fd73e0bf8cf2f5a0.mp3" }, { "key": "original-vn_S_ASMR_reg-nl-0-seed1701-scale0.5", "bucket": "vn_S_ASMR_reg", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De deur staat nog open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.5, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "7fdadb2748e4ce9d30fd95df589180b3a7265594f2819691a64304e7bc344e17", "task_id": "vn_S_ASMR_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De deur staat nog open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4846596717834473, "target_z": -0.08203127296698219, "ce": 6.1020402908325195, "genuineness": 4.469690799713135, "identity": 0.6472762823104858, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/nl/554c8022ad4bbe8bfd21f814.mp3" }, { "key": "original-vn_S_ASMR_reg-nl-0-seed1701-scale0.75", "bucket": "vn_S_ASMR_reg", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De deur staat nog open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.75, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "875cc9636aaf880009bd2d8b0e492d4c6611084ddc4da40c98f9204b1420040d", "task_id": "vn_S_ASMR_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De deur staat nog open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6498897075653076, "target_z": -0.6874999783840168, "ce": 6.078212261199951, "genuineness": 4.256305694580078, "identity": 0.6262050271034241, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd, zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/nl/684d0ba43adc667be8f7d245.mp3" }, { "key": "original-vn_S_ASMR_reg-nl-0-seed1701-scale1.0", "bucket": "vn_S_ASMR_reg", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De deur staat nog open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 1.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "a60c99e768ce06d637d95f790ad7fb984ec9b10ade37893971ad0b3ea335be4a", "task_id": "vn_S_ASMR_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De deur staat nog open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6391184329986572, "target_z": -0.6953125135099896, "ce": 5.967014789581299, "genuineness": 4.2249250411987305, "identity": 0.4945828318595886, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/nl/34b2118486a0627081fa41a4.mp3" }, { "key": "original-vn_S_ASMR_reg-nl-0-seed1702-scale0.0", "bucket": "vn_S_ASMR_reg", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De deur staat nog open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "79d7a94c05d0dd7dc43dc6f410726c001550a8a3a110c0f6e7f4b4e0330d564e", "task_id": "vn_S_ASMR_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De deur staat nog open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8545429706573486, "target_z": -0.539062502701998, "ce": 6.213237762451172, "genuineness": 4.482243061065674, "identity": 0.6472454071044922, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/nl/10855af39427b8b8773fd14b.mp3" }, { "key": "original-vn_S_ASMR_reg-nl-0-seed1702-scale0.5", "bucket": "vn_S_ASMR_reg", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De deur staat nog open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.5, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "1602f4d7e03a94cd83969663558b8793855b293e84bd5bf4a4f3226435a21d06", "task_id": "vn_S_ASMR_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De deur staat nog open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.215378999710083, "target_z": -0.2773437864769717, "ce": 6.205295085906982, "genuineness": 4.457139015197754, "identity": 0.6198285818099976, "hypothesis": "de deur staat nog open ik heb je jas op de stoel gelegd zeg maar wanneer je klaar bent", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/nl/95ffcd4a9e807a324db9580b.mp3" }, { "key": "original-vn_S_ASMR_reg-nl-0-seed1702-scale0.75", "bucket": "vn_S_ASMR_reg", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De deur staat nog open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.75, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "aa2ee1a338057255732482a39e5fcf4124818ff09545b70cc20146b0cb6a65a7", "task_id": "vn_S_ASMR_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De deur staat nog open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8383861184120178, "target_z": -0.5507812621589906, "ce": 6.09409761428833, "genuineness": 4.369274139404297, "identity": 0.7328923344612122, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/nl/4b6784197ef356351243e016.mp3" }, { "key": "original-vn_S_ASMR_reg-nl-0-seed1702-scale1.0", "bucket": "vn_S_ASMR_reg", "language": "nl", "sentences": [ "De deur staat nog open.", "Ik heb je jas op de stoel gelegd.", "Zeg maar wanneer je klaar bent." ], "text": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De deur staat nog open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 1.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "e373dace8e07d88981ea004a3793c4589b6b94e79410d687abe78624bf4bc854", "task_id": "vn_S_ASMR_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] De deur staat nog open.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.806072473526001, "target_z": -0.5742187378410094, "ce": 6.205295085906982, "genuineness": 4.432034492492676, "identity": 0.6219120025634766, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/nl/b84b197c19c84fbaa383a085.mp3" }, { "key": "original-vn_S_ASMR_reg-nl-1-seed1701-scale0.0", "bucket": "vn_S_ASMR_reg", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond naast het oude raam.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "ea48e6724c756270ca65951cf2369abe14179b1ffc2807debb70b9d6f8a54198", "task_id": "vn_S_ASMR_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond naast het oude raam.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1615228652954102, "target_z": -0.3164062891789696, "ce": 6.070269584655762, "genuineness": 4.256305694580078, "identity": 0.4078293740749359, "hypothesis": "ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/nl/de8a04232e6da77b70d9371c.mp3" }, { "key": "original-vn_S_ASMR_reg-nl-1-seed1701-scale0.5", "bucket": "vn_S_ASMR_reg", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond naast het oude raam.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.5, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "e14985ccba0851a49b680b6c263c200da1a9ff92db0ef5b2f423dd7877766fc6", "task_id": "vn_S_ASMR_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond naast het oude raam.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7252882719039917, "target_z": -0.6328124918940062, "ce": 6.284721851348877, "genuineness": 4.262581825256348, "identity": 0.6878845691680908, "hypothesis": "ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/nl/0bf00e02de3b40f3068ef3fc.mp3" }, { "key": "original-vn_S_ASMR_reg-nl-1-seed1701-scale0.75", "bucket": "vn_S_ASMR_reg", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond naast het oude raam.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.75, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "1b1990169181644adb87fadd65d349499f00a4373b1929ccbaeda683d8065c49", "task_id": "vn_S_ASMR_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond naast het oude raam.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4267643690109253, "target_z": -0.1240234331092534, "ce": 6.1894097328186035, "genuineness": 4.369274139404297, "identity": 0.5842045545578003, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/nl/dec1fbadf0c90e5386ec3007.mp3" }, { "key": "original-vn_S_ASMR_reg-nl-1-seed1701-scale1.0", "bucket": "vn_S_ASMR_reg", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond naast het oude raam.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 1.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "3ead19b7e451b7a78c90bcc62cfff7dacd86cfe1a9c1d5cf307e1a8c3f0afed6", "task_id": "vn_S_ASMR_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond naast het oude raam.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9353271722793579, "target_z": -0.480468748649001, "ce": 6.197352409362793, "genuineness": 4.118232727050781, "identity": 0.7949340343475342, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/nl/b86f2a39eb79d2f92147a94a.mp3" }, { "key": "original-vn_S_ASMR_reg-nl-1-seed1702-scale0.0", "bucket": "vn_S_ASMR_reg", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond naast het oude raam.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "1abd4cca40f82ebd74d521f38bab4ba9a751e44b134619cf1b12e424a99604da", "task_id": "vn_S_ASMR_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond naast het oude raam.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6983602046966553, "target_z": -0.6523437432450052, "ce": 6.23706579208374, "genuineness": 4.293961524963379, "identity": 0.6377138495445251, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het Oude Raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/nl/e102ba67aeaee7495f3ff53c.mp3" }, { "key": "original-vn_S_ASMR_reg-nl-1-seed1702-scale0.5", "bucket": "vn_S_ASMR_reg", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond naast het oude raam.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.5, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "c6dc1e3e6c04dd2db9a594ce94d12da840a34284a31ce1446b135bf34e7365f5", "task_id": "vn_S_ASMR_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond naast het oude raam.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9083991050720215, "target_z": -0.5, "ce": 6.197352409362793, "genuineness": 4.344170093536377, "identity": 0.6787704825401306, "hypothesis": "ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/nl/ef03cf0a30095fd58007e531.mp3" }, { "key": "original-vn_S_ASMR_reg-nl-1-seed1702-scale0.75", "bucket": "vn_S_ASMR_reg", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond naast het oude raam.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.75, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "9f18a25ee62d48a957bfd1e92970377b82256b38673f843a1ba2cd1a881c3d1c", "task_id": "vn_S_ASMR_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond naast het oude raam.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9083991050720215, "target_z": -0.5, "ce": 6.149696350097656, "genuineness": 4.1998209953308105, "identity": 0.6241933107376099, "hypothesis": "ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/nl/764ecbda2d3173a63b06e62e.mp3" }, { "key": "original-vn_S_ASMR_reg-nl-1-seed1702-scale1.0", "bucket": "vn_S_ASMR_reg", "language": "nl", "sentences": [ "Ik heb het kleine doosje gevonden.", "Het stond naast het oude raam.", "We kunnen er samen in kijken." ], "text": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond naast het oude raam.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "directions": [ "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 1.0, "frames_generated": 128, "seconds_generated": 10.24, "ended": true, "contract_sha256": "96c7769c29fcf9e401567aff105ac59fd323f2e458b19f42ff56d7cdf9fd59a1", "mp3_sha256": "f9e59e6a4380e49021153699205741d927a45ea85c29dd079a2a0fa86e93ccbe", "task_id": "vn_S_ASMR_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.8 seconds duration] Het stond naast het oude raam.\n(Strong ASMR Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7737587690353394, "target_z": -0.5976562567549948, "ce": 6.149696350097656, "genuineness": 4.262581825256348, "identity": 0.5813576579093933, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/audio/vn_S_ASMR_reg/nl/833b571abe78c7371a822f16.mp3" } ], "source_standard_audio_unchanged": true }