{ "complete": true, "report": { "bucket": "vn_S_DRAM_reg", "label": "Strong Dramatic Style", "definition": { "id": "vn_S_DRAM_reg", "category": "talking_style", "head": "vn_S_DRAM_reg", "pole": "high", "label": "Strong Dramatic Style", "floor": 4.0, "canonical_dimension_name": "Dramatic Style" }, "recipe_status": "direct_manual_recipe", "manual_provenance": [ { "id": "rank4/style_s_dram/full-v2-sealed/sealed-selected", "mission": "style_s_dram", "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": 6.101563796997071, "genuineness": 4.276514358520508, "wer": 0.014331759178533372, "identity": 0.5384404063224792, "voiceclap": 0.03302472951821983, "fitness": 0.7192909999247086, "steering_v2": 0.7989120904049619, "raw_target": 3.845650939941406, "target_cdf": 0.7989120904049619, "family_hit_all": null, "nominal_hit_all": null, "n": 100, "valid_fraction": 1, "by_arm": { "reference": { "n": 50, "ce": 6.093541707992554, "genuineness": 4.295091404914856, "wer": 0.01430718621847654, "identity": 0.5384404063224792, "fitness": 0.7154646424113221, "steering_v2": 0.7898683149935523 }, "none": { "n": 50, "ce": 6.109585886001587, "genuineness": 4.25793731212616, "wer": 0.014356332138590203, "identity": null, "fitness": 0.7231173574380949, "steering_v2": 0.8079558658163715 } } } } ], "selected_wording": { "id": "manual-21a285777a88", "general": "Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.", "inline": [ "Initiate at the very edge of audibility, focusing on breath control for intimate delivery.", "Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.", "Hold the peak energy level briefly, then allow a controlled, slow dissipation of power." ], "cfg": 3, "recipe": "P06", "strategy": "local-Gemma-mutation", "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": 0.8802734446526167, "target_raw": 3.8089347958564757, "ce": 5.838542222976685, "genuineness": 4.36864664554596, "identity": 0.4792092226445675, "wer": 0.027339181286549708, "valid_fraction": 0.9, "identity_delta": -0.04566736593842506, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, "manual_adapter": { "config": { "scale": 0.5, "cfg": 4.0 }, "n": 20, "target_z": 1.0431640615599418, "target_raw": 3.9945182681083677, "ce": 5.6378905296325685, "genuineness": 4.234025633335113, "identity": 0.4916503380984068, "wer": 0.04513888888888889, "valid_fraction": 0.95, "identity_delta": -0.03322625048458576, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, "grid": [ { "config": { "scale": 0.0, "cfg": 4.0 }, "n": 20, "target_z": 0.8802734446526167, "target_raw": 3.8089347958564757, "ce": 5.838542222976685, "genuineness": 4.36864664554596, "identity": 0.4792092226445675, "wer": 0.027339181286549708, "valid_fraction": 0.9, "identity_delta": -0.04566736593842506, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 3.0 }, "n": 20, "target_z": 0.6576660304309166, "target_raw": 3.555315172672272, "ce": 5.998189926147461, "genuineness": 4.446783351898193, "identity": 0.5141654938459397, "wer": 0.0, "valid_fraction": 1, "identity_delta": -0.010711094737052918, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 2.5 }, "n": 20, "target_z": 0.408447269477195, "target_raw": 3.271376883983612, "ce": 6.129045391082764, "genuineness": 4.516761112213135, "identity": 0.5516492769122123, "wer": 0.02222222222222222, "valid_fraction": 0.95, "identity_delta": 0.026772688329219817, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 2.0 }, "n": 20, "target_z": 0.16123047477863411, "target_raw": 2.9897194623947145, "ce": 6.1957638502120975, "genuineness": 4.507347130775452, "identity": 0.5504604265093803, "wer": 0.021052631578947368, "valid_fraction": 0.95, "identity_delta": 0.025583837926387788, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 1.25 }, "n": 20, "target_z": 0.021612546867541663, "target_raw": 2.830650877952576, "ce": 6.246199822425842, "genuineness": 4.412892758846283, "identity": 0.5797602608799934, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.054883672297000884, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 1.5 }, "n": 20, "target_z": 0.01778564219387703, "target_raw": 2.8262908339500425, "ce": 6.212840628623963, "genuineness": 4.499188256263733, "identity": 0.5108924180269241, "wer": 0.0, "valid_fraction": 1, "identity_delta": -0.013984170556068421, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 1.0 }, "n": 20, "target_z": -0.24147947726407606, "target_raw": 2.530906593799591, "ce": 6.220783281326294, "genuineness": 4.419796335697174, "identity": 0.5361567333340644, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.01128014475107193, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 5.0 }, "n": 20, "target_z": 1.1615234564237804, "target_raw": 4.129366719722748, "ce": 5.670505118370056, "genuineness": 4.196369290351868, "identity": 0.4611756257712841, "wer": 0.0, "valid_fraction": 1, "identity_delta": -0.06370096281170845, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 4.0 }, "n": 20, "target_z": 1.0431640615599418, "target_raw": 3.9945182681083677, "ce": 5.6378905296325685, "genuineness": 4.234025633335113, "identity": 0.4916503380984068, "wer": 0.04513888888888889, "valid_fraction": 0.95, "identity_delta": -0.03322625048458576, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 3.0 }, "n": 20, "target_z": 1.0376953045299413, "target_raw": 3.98828763961792, "ce": 5.843704962730408, "genuineness": 4.352956557273865, "identity": 0.5142975568771362, "wer": 0.0027777777777777775, "valid_fraction": 1, "identity_delta": -0.010579031705856324, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 2.5 }, "n": 20, "target_z": 0.8062378168457174, "target_raw": 3.724585008621216, "ce": 5.859987425804138, "genuineness": 4.254736483097076, "identity": 0.47739173769950866, "wer": 0.009236111111111112, "valid_fraction": 1, "identity_delta": -0.047484850883483885, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 2.5 }, "n": 20, "target_z": 0.6646362333783323, "target_raw": 3.5632564187049867, "ce": 5.952916669845581, "genuineness": 4.3429148435592655, "identity": 0.5242415636777877, "wer": 0.008680555555555556, "valid_fraction": 1, "identity_delta": -0.000635024905204773, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 2.5 }, "n": 20, "target_z": 0.6427246152503637, "target_raw": 3.5382922172546385, "ce": 6.031946206092835, "genuineness": 4.445528125762939, "identity": 0.512395977973938, "wer": 0.010526315789473684, "valid_fraction": 0.95, "identity_delta": -0.012480610609054565, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 2.0 }, "n": 20, "target_z": 0.39423828225136637, "target_raw": 3.2551883935928343, "ce": 6.1044231176376345, "genuineness": 4.448352336883545, "identity": 0.5540544293820858, "wer": 0.008204334365325076, "valid_fraction": 1, "identity_delta": 0.029177840799093246, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 2.0 }, "n": 20, "target_z": 0.3860351519379939, "target_raw": 3.245842456817627, "ce": 6.1175285339355465, "genuineness": 4.41916868686676, "identity": 0.53771997615695, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.012843387573957444, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 2.0 }, "n": 20, "target_z": 0.33706056153786434, "target_raw": 3.1900450468063353, "ce": 6.055575680732727, "genuineness": 4.395319736003875, "identity": 0.5468667283654213, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.021990139782428742, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 1.5 }, "n": 20, "target_z": 0.31162109953340156, "target_raw": 3.1610615253448486, "ce": 6.166375970840454, "genuineness": 4.398771631717682, "identity": 0.5506157904863358, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.025739201903343202, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 1.5 }, "n": 20, "target_z": 0.2304504201729911, "target_raw": 3.0685826778411864, "ce": 6.159624719619751, "genuineness": 4.392181766033173, "identity": 0.5684403970837593, "wer": 0.005409356725146199, "valid_fraction": 1, "identity_delta": 0.04356380850076676, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 1.25 }, "n": 20, "target_z": 0.10372087003233356, "target_raw": 2.924197995662689, "ce": 6.186629819869995, "genuineness": 4.350446140766143, "identity": 0.53172476477921, "wer": 0.047058823529411764, "valid_fraction": 0.95, "identity_delta": 0.006848176196217537, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 1.5 }, "n": 20, "target_z": 0.06564944282112832, "target_raw": 2.880822706222534, "ce": 6.192586779594421, "genuineness": 4.384964334964752, "identity": 0.5632218480110168, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.03834525942802429, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 1.25 }, "n": 20, "target_z": 0.0035888865148922555, "target_raw": 2.8101162791252134, "ce": 6.195763897895813, "genuineness": 4.380571174621582, "identity": 0.5526605062186718, "wer": 0.006274509803921568, "valid_fraction": 1, "identity_delta": 0.027783917635679244, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 1.0 }, "n": 20, "target_z": -0.012951670008979567, "target_raw": 2.7912714004516603, "ce": 6.1826584815979, "genuineness": 4.342287254333496, "identity": 0.5757212847471237, "wer": 0.0027777777777777775, "valid_fraction": 1, "identity_delta": 0.05084469616413116, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 1.25 }, "n": 20, "target_z": -0.025793444434981015, "target_raw": 2.7766405940055847, "ce": 6.164390325546265, "genuineness": 4.355153155326843, "identity": 0.5534647554159164, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.028588166832923888, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 1.0 }, "n": 20, "target_z": -0.04908752032749156, "target_raw": 2.7501013398170473, "ce": 6.214429116249084, "genuineness": 4.344797730445862, "identity": 0.5143937900662422, "wer": 0.015789473684210527, "valid_fraction": 0.95, "identity_delta": -0.010482798516750335, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 1.0 }, "n": 20, "target_z": -0.05292966996587095, "target_raw": 2.7457239270210265, "ce": 6.18464412689209, "genuineness": 4.322831535339356, "identity": 0.5883373498916626, "wer": 0.0027777777777777775, "valid_fraction": 1, "identity_delta": 0.06346076130867004, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 5.0 }, "n": 20, "target_z": 1.397656241008139, "target_raw": 4.39839597940445, "ce": 5.461166024208069, "genuineness": 4.004636287689209, "identity": 0.4300206571817398, "wer": 0.04429824561403509, "valid_fraction": 0.95, "identity_delta": -0.09485593140125274, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 5.0 }, "n": 20, "target_z": 1.1784179116721183, "target_raw": 4.1486147999763485, "ce": 5.3836256980896, "genuineness": 4.056570518016815, "identity": 0.4237039387226105, "wer": 0.005409356725146199, "valid_fraction": 1, "identity_delta": -0.10117264986038207, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 5.0 }, "n": 20, "target_z": 1.144531232506743, "target_raw": 4.110007250308991, "ce": 5.560548710823059, "genuineness": 4.156202661991119, "identity": 0.45771211832761766, "wer": 0.0027777777777777775, "valid_fraction": 1, "identity_delta": -0.0671644702553749, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 4.0 }, "n": 20, "target_z": 1.1142578401704089, "target_raw": 4.0755163669586185, "ce": 5.606417632102966, "genuineness": 4.157457840442658, "identity": 0.44518798887729644, "wer": 0.00804093567251462, "valid_fraction": 1, "identity_delta": -0.07968859970569611, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 4.0 }, "n": 20, "target_z": 1.090234370667558, "target_raw": 4.048146104812622, "ce": 5.530515503883362, "genuineness": 4.155888891220092, "identity": 0.4431240729987621, "wer": 0.008534356725146198, "valid_fraction": 1, "identity_delta": -0.08175251558423043, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 3.0 }, "n": 20, "target_z": 1.0171875101362797, "target_raw": 3.964922833442688, "ce": 5.700538277626038, "genuineness": 4.202645373344422, "identity": 0.4296868611127138, "wer": 0.008350533195734434, "valid_fraction": 1, "identity_delta": -0.09518972747027873, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 3.0 }, "n": 20, "target_z": 0.7645996119805961, "target_raw": 3.67714604139328, "ce": 5.949441742897034, "genuineness": 4.384336757659912, "identity": 0.44891746193170545, "wer": 0.0027777777777777775, "valid_fraction": 1, "identity_delta": -0.07595912665128708, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" } ], "development_standard": [ { "scale": 0.0, "n": 20, "target_z": -0.00358887605163576, "target_raw": 2.801938557624817, "ce": 6.19933807849884, "genuineness": 4.257246935367585, "identity": 0.5248765885829926, "wer": 0.013157894736842105, "valid_fraction": 0.95, "identity_delta": 0.0, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "scale": 0.5, "n": 20, "target_z": -0.014587422286778194, "target_raw": 2.789407765865326, "ce": 6.1806727886199955, "genuineness": 4.246577668190002, "identity": 0.5971888802945614, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.07231229171156883, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "scale": 0.75, "n": 20, "target_z": -0.18460082781100093, "target_raw": 2.595709204673767, "ce": 6.181069922447205, "genuineness": 4.203900599479676, "identity": 0.5892029970884323, "wer": 0.002631578947368421, "valid_fraction": 1, "identity_delta": 0.06432640850543976, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "scale": 1.0, "n": 20, "target_z": -0.06030276060466741, "target_raw": 2.7373236656188964, "ce": 6.179481363296508, "genuineness": 4.258188366889954, "identity": 0.587454704940319, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.0625781163573265, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" } ] }, "heldout": { "standard": { "n": 50, "target_z": -0.3134814414265855, "target_raw": 2.4488737869262693, "ce": 6.255175065994263, "genuineness": 4.342161731719971, "identity": 0.5841461846232414, "wer": 0.00927159796725014, "valid_fraction": 1, "identity_delta": 0.0, "identity_pairs": 50, "gates_pass": true, "gate_scope": "heldout_50_reference_pairs" }, "manual": { "n": 50, "target_z": 0.992753914461204, "target_raw": 3.9370853090286255, "ce": 5.76761414527893, "genuineness": 4.151809430122375, "identity": 0.46429080963134767, "wer": 0.015734236777715037, "valid_fraction": 1, "identity_delta": -0.11985537499189376, "identity_pairs": 50, "gates_pass": false, "gate_scope": "heldout_50_reference_pairs" }, "adapter": { "n": 50, "target_z": 1.0135937372806039, "target_raw": 3.960828399658203, "ce": 5.682071561813355, "genuineness": 4.116600837707519, "identity": 0.4543541517108679, "wer": 0.02412020168999574, "valid_fraction": 0.98, "identity_delta": -0.12979203291237354, "identity_pairs": 50, "gates_pass": false, "gate_scope": "heldout_50_reference_pairs" }, "matched_no_adapter": { "n": 50, "target_z": 0.992753914461204, "target_raw": 3.9370853090286255, "ce": 5.76761414527893, "genuineness": 4.151809430122375, "identity": 0.46429080963134767, "wer": 0.015734236777715037, "valid_fraction": 1, "identity_delta": -0.11985537499189376, "identity_pairs": 50, "gates_pass": false, "gate_scope": "heldout_50_reference_pairs" } }, "adapter_vs_matched_cfg": { "n": 50, "delta_z": 0.020839822819399806, "ci95_z": [ -0.09390625801093073, 0.14990232131530667 ], "language_deltas": { "en": -0.17109374722069745, "de": 0.24462888607962077, "fr": -0.013085976369362973, "nl": 0.03593747743860317, "es": 0.0078124741688355035 }, "languages_improved": 3, "clear_gain": false, "delta_ce": -0.08554258346557617, "delta_genuineness": -0.03520859241485596, "delta_wer": 0.008385964912280702, "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": 0.020839822819399806, "ci95_z": [ -0.09390625801093073, 0.14990232131530667 ], "language_deltas": { "en": -0.17109374722069745, "de": 0.24462888607962077, "fr": -0.013085976369362973, "nl": 0.03593747743860317, "es": 0.0078124741688355035 }, "languages_improved": 3, "clear_gain": false, "delta_ce": -0.08554258346557617, "delta_genuineness": -0.03520859241485596, "delta_wer": 0.008385964912280702, "note": "Exploratory language-cluster bootstrap; five voices, one fixed reference voice per language. No familywise claim across72targets." }, "manual_vs_standard": { "n": 50, "delta_z": 1.3062353558877895, "ci95_z": [ 1.1866552756680946, 1.4258203203556168 ], "language_deltas": { "en": 1.4957031276975583, "de": 1.1784667646166984, "fr": 1.2150146684392817, "nl": 1.4613281283515118, "es": 1.1806640903338972 }, "languages_improved": 5, "clear_gain": true, "delta_ce": -0.487560920715332, "delta_genuineness": -0.19035230159759522, "delta_wer": 0.006462638810464897, "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": 1793023652.0014107 }, "rows": [ { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "00a2cff526f2d1d9af341b2de85c640085a6ef8d830d709cfa0dd94fa15d7d87", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "9d788147729658b41766b9634a566ff88821de4a4c6d9a3bbc75ee416da363c4", "generation_seconds_batch": 21.56047468702309, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.603532314300537, "target_z": -0.17773431573546125, "ce": 6.086154937744141, "genuineness": 4.369274139404297, "identity": 0.659622073173523, "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_DRAM_reg/00a2cff526f2d1d9af341b2de85c640085a6ef8d830d709cfa0dd94fa15d7d87.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "00d27b09eff64968245b262b8999230f12ea44b82016dae1eefc6aa6061459ed", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "cae15b3705c1cf37c648776b53edf8fdc359bf3f50378d306d66260c6d267145", "generation_seconds_batch": 15.174126013065688, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9809443950653076, "target_z": 1.0312499803813941, "ce": 6.165581703186035, "genuineness": 4.21237325668335, "identity": 0.3009568452835083, "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_DRAM_reg/00d27b09eff64968245b262b8999230f12ea44b82016dae1eefc6aa6061459ed.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "01bbbe0b53effd095231ca14520e02e99619dc8ecd6052385422d3a9772f3470", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "86f4c7e0fd708df046ebf44c419df6a79d7dd58d57c892293e1576d40d24d009", "generation_seconds_batch": 21.200859484961256, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.866386651992798, "target_z": 0.05297860661650303, "ce": 6.24500846862793, "genuineness": 4.570107460021973, "identity": 0.570807933807373, "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_DRAM_reg/01bbbe0b53effd095231ca14520e02e99619dc8ecd6052385422d3a9772f3470.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "01e1924d01df6110a0c90292d6e848fd1d13d837d41be8cb529df780c177d29c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "11fd1b16744b669ee6b774222cf2d1556d8cb77dcd4b12776f64191e991354dd", "generation_seconds_batch": 24.32487613603007, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4468913078308105, "target_z": 0.5625000653953531, "ce": 6.078212261199951, "genuineness": 4.632867813110352, "identity": 0.47090479731559753, "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_DRAM_reg/01e1924d01df6110a0c90292d6e848fd1d13d837d41be8cb529df780c177d29c.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "28993da5c8a37783baec96a70f1738289c71e1ceb5046749841478827c50a18f", "generation_seconds_batch": 16.003403531969525, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.945103645324707, "target_z": 0.12207020970667934, "ce": 6.070269584655762, "genuineness": 4.2751336097717285, "identity": 0.8069926500320435, "hypothesis": "J'ai écrit le numéro sur l'enveloppe, il est bien, et vérifie les deux derniers chiffres. 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_DRAM_reg/01f5d5f22d9b51f4e674eec92be80d0710e62f5704093e3409bc811236eec3e5.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "0200f9d2406cec63b5a17ba088932b225e5f994b326a9f0414370120059992fb", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "54515300bdea30d8e6fd419fe0e1853d8d1b7de487b8579d63860179398b0ac8", "generation_seconds_batch": 21.37001985195093, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2466213703155518, "target_z": 0.3867188211174465, "ce": 6.078212261199951, "genuineness": 4.6705241203308105, "identity": 0.7481440305709839, "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_DRAM_reg/0200f9d2406cec63b5a17ba088932b225e5f994b326a9f0414370120059992fb.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "026b44ce9842bb53ecc66da1171a2de5bb96770f3c861a6ebe301a10c6503dc5", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "ff82c1199302fd6238faf677f6b6c9d8e6629678b1b7caaeecfc6f1fb21f5373", "generation_seconds_batch": 20.75173568201717, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1553871631622314, "target_z": 0.30664058944127676, "ce": 6.117925643920898, "genuineness": 4.356721878051758, "identity": 0.6831796169281006, "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_DRAM_reg/026b44ce9842bb53ecc66da1171a2de5bb96770f3c861a6ebe301a10c6503dc5.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "027edff44aed68b3687e7df7eb77bc98fd8f9f595f3e35267e2f2f0fe108133c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "52662a708ab8d2462cf28e4f44181e0b4ff0768b48168c4fc6d2aecbdc0972ba", "generation_seconds_batch": 20.858004731009714, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.162062883377075, "target_z": 0.3125000130790706, "ce": 5.77639102935791, "genuineness": 4.017816066741943, "identity": 0.496050626039505, "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_DRAM_reg/027edff44aed68b3687e7df7eb77bc98fd8f9f595f3e35267e2f2f0fe108133c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "0284abae80b9cd4fe9c810410c7aa081bd474d7428f76ee58e33f4f99e368e0c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "8d36469ac8f87e4a88f3eea6be149f5d722f6aac5b953f77e7f4f3131ef4b083", "generation_seconds_batch": 20.608124705962837, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.052151679992676, "target_z": 1.0937501504093121, "ce": 6.133810997009277, "genuineness": 3.961331605911255, "identity": 0.48126181960105896, "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_DRAM_reg/0284abae80b9cd4fe9c810410c7aa081bd474d7428f76ee58e33f4f99e368e0c.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": "02c6b071223c34a180b10c66ee373672d8053ae1c23fecda022034d1565ea8e7", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "b69b4b5da7dbfedcb39472c91bc82b5b16e0e8f69a2edfa5fa543d516e222973", "generation_seconds_batch": 15.383068852010183, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.141160488128662, "target_z": 1.1718751013627973, "ce": 5.6532793045043945, "genuineness": 4.262581825256348, "identity": 0.25758644938468933, "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_DRAM_reg/02c6b071223c34a180b10c66ee373672d8053ae1c23fecda022034d1565ea8e7.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "02ceba7a7c0470793b5050c76e7240750e8d76d8ca9efc6dc59f2c2b45f266d4", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "ec32514e70c67f2732a985d541b0c3a1cbcd39c99c2f1c7cf5b2883398d8c206", "generation_seconds_batch": 23.789757214020938, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.977369546890259, "target_z": 0.1503906875343064, "ce": 6.0305562019348145, "genuineness": 4.193545341491699, "identity": 0.5022362470626831, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/02ceba7a7c0470793b5050c76e7240750e8d76d8ca9efc6dc59f2c2b45f266d4.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "035125cb0db23bdac800bae6bca0c2ffe69a318d860b901379b89383cf3fb9ae", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "27f0b30bfbb9c11a60fbe94e55ca75d7adef18fe3c6353891b4ff65696d531f0", "generation_seconds_batch": 20.895830499008298, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9417660236358643, "target_z": 0.11914070715291233, "ce": 6.117925643920898, "genuineness": 4.362998008728027, "identity": 0.7051109075546265, "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_DRAM_reg/035125cb0db23bdac800bae6bca0c2ffe69a318d860b901379b89383cf3fb9ae.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "038696536ef7464041b18419f079ca3c5400918da347a9a2cbaf6e5f38977cd2", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "077daff7e1f5332ba9660e94b2543168a89a3ccdaaa257701aacbf03e3f26d81", "generation_seconds_batch": 20.61788582499139, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.455792188644409, "target_z": 0.5703125604907017, "ce": 5.816103935241699, "genuineness": 4.570107460021973, "identity": 0.7727259397506714, "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_DRAM_reg/038696536ef7464041b18419f079ca3c5400918da347a9a2cbaf6e5f38977cd2.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "03da1ebb831c1950e3e342cc255f2f039c5e11b3e65dd22b8a33621f1565bd0a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "9cfaa4eddd737b3c99c280dc095bc30390b0be8e94ceff35adbdad8e1eca5b07", "generation_seconds_batch": 14.236820242018439, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.536775588989258, "target_z": -0.2363281335831401, "ce": 6.459460258483887, "genuineness": 4.432034492492676, "identity": 0.4726344347000122, "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_DRAM_reg/03da1ebb831c1950e3e342cc255f2f039c5e11b3e65dd22b8a33621f1565bd0a.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "043ed2c09a684fed0fc9f3c965aab73f643c2e435011068289199949579d8746", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "390582d9563e10827d36f0ba0d67e4e9a7c3ede9020d06ee9f845861a63fa2c2", "generation_seconds_batch": 20.54670853400603, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.315603256225586, "target_z": 0.44726571042268, "ce": 6.316492557525635, "genuineness": 4.570107460021973, "identity": 0.6125780344009399, "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_DRAM_reg/043ed2c09a684fed0fc9f3c965aab73f643c2e435011068289199949579d8746.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "054e3f2430b13e90854f8bf415ffa1ab76e354171c789cfeffa40c6a45076b45", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "46834927cd12025d7d247bcd09e0f5929924527e561add14da7b3b1171e77b6f", "generation_seconds_batch": 21.278364436002448, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.9244384765625, "target_z": 1.8593750882837268, "ce": 6.046441555023193, "genuineness": 3.672633647918701, "identity": 0.40630143880844116, "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_DRAM_reg/054e3f2430b13e90854f8bf415ffa1ab76e354171c789cfeffa40c6a45076b45.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "05573dc93c396f430e46f7cffe9b3a443da608e051ca2f909c9e79e7fcecce9c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "32664dba9b479201446d1ea20c751cbc5adfbe279fb75fedfd1b6eb564a0d726", "generation_seconds_batch": 20.3628170019947, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0096352100372314, "target_z": 0.17871095609680354, "ce": 5.998785495758057, "genuineness": 4.6956281661987305, "identity": 0.6370752453804016, "hypothesis": "Ich hab die kleine Schachtel 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_DRAM_reg/05573dc93c396f430e46f7cffe9b3a443da608e051ca2f909c9e79e7fcecce9c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "0578aaf4b3535782b5b6507d2311fd27d7981c8dbf9fc2f7359ec56351290531", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "497f80859016022c37d260bec65d1c1b1ec3c41118ede6cadb8e6f3ad66ed82e", "generation_seconds_batch": 14.750742144999094, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5002965927124023, "target_z": 0.6093750359674442, "ce": 6.078212261199951, "genuineness": 4.394378662109375, "identity": 0.4474450349807739, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/0578aaf4b3535782b5b6507d2311fd27d7981c8dbf9fc2f7359ec56351290531.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "bc402727a5535427b2d9e9d2d4969a7e582736e64f9c76f98f745faa445e7e62", "generation_seconds_batch": 15.506583926035091, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7910072803497314, "target_z": -0.013183493919906282, "ce": 6.292664527893066, "genuineness": 4.519899368286133, "identity": 0.6254240274429321, "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_DRAM_reg/059867605942d1f4c46c2c44d936e08b7fbe89dd4feef4c05500a7ed426c27a3.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "06782dfdbc6eae417445db027d29371f3d162d0d3f23a2aa23698279e3b4df6d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "5d993cd5e154557a8acc7d2b43b5780f6543fba109ea0d964e3a8b2e6a64a0b2", "generation_seconds_batch": 21.03318190993741, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8118276596069336, "target_z": 0.8828125735697723, "ce": 5.474569320678711, "genuineness": 4.2374773025512695, "identity": 0.38483572006225586, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/06782dfdbc6eae417445db027d29371f3d162d0d3f23a2aa23698279e3b4df6d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "068134a4e86fb9ba83a501b4376f7031aee34f4f2d2a68a25dfb99bd15f3f267", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "7a30ec1dd2813184c0f69b061f473eb5bbe41436bf372a0628139d7fdd60d7ca", "generation_seconds_batch": 20.410747074056417, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.478044271469116, "target_z": 0.5898436935965079, "ce": 6.109982967376709, "genuineness": 4.570107460021973, "identity": 0.6754924654960632, "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_DRAM_reg/068134a4e86fb9ba83a501b4376f7031aee34f4f2d2a68a25dfb99bd15f3f267.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": "06ab81c0cdfe11cb464975bb30e0b224b2315a42ad776996b68c0da3deffea5f", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "640612ee21d743fcb2b9bd3a9d41d24407d2bfac2f36ed39b4f726b232f664dd", "generation_seconds_batch": 15.638461338938214, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.105556964874268, "target_z": 1.1406251209814033, "ce": 5.069493293762207, "genuineness": 3.9048471450805664, "identity": 0.34255877137184143, "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_DRAM_reg/06ab81c0cdfe11cb464975bb30e0b224b2315a42ad776996b68c0da3deffea5f.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "07a1018bc261aacf722f298c4963860f37d8ce214224f5d8b1268b86017dad7c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "1979b65404c0de4cab0ac428428230fff5cd3cac7c14a6111a36e1ff5aa6f0d0", "generation_seconds_batch": 21.96490627503954, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.90083646774292, "target_z": 0.9609375245232574, "ce": 5.935244083404541, "genuineness": 4.482243061065674, "identity": 0.3498307168483734, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/07a1018bc261aacf722f298c4963860f37d8ce214224f5d8b1268b86017dad7c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "07bdc1b481189dc94c7eb2ee9ba6ebba49b5fc62fce8d16f65c45f07f11c343d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "3103aaa2b30cdda1049d7767484ff4e468a826f391dadb580a8f44d73179a98a", "generation_seconds_batch": 16.426489592064172, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7673230171203613, "target_z": 0.8437498888278997, "ce": 5.863759994506836, "genuineness": 4.3504462242126465, "identity": 0.371840238571167, "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_DRAM_reg/07bdc1b481189dc94c7eb2ee9ba6ebba49b5fc62fce8d16f65c45f07f11c343d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "07c52701f530edb1fc49f0d2a2cf585947c2aeebdf7895fd628aa606e50a9ece", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "6f86f48d158dd1f3d2b1fa5f27877ad566ad0abca1682d9f57c6dd037a0f9a1a", "generation_seconds_batch": 16.126302730990574, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2555222511291504, "target_z": 0.394531316212795, "ce": 6.046441555023193, "genuineness": 4.532451152801514, "identity": 0.4711742103099823, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/07c52701f530edb1fc49f0d2a2cf585947c2aeebdf7895fd628aa606e50a9ece.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "07cb4f4452986630730a5a3a9d1eb5a1d494adb928992b235122f7e01b6c0e5d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "668bf10a4df8d4fa54d032f7921fe03607921c2887906a169ecf784d516e1d3d", "generation_seconds_batch": 21.247972128912807, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.256872177124023, "target_z": 1.273437746867458, "ce": 5.518253803253174, "genuineness": 4.055472373962402, "identity": 0.770571231842041, "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": 0.03999999999999915, "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_DRAM_reg/07cb4f4452986630730a5a3a9d1eb5a1d494adb928992b235122f7e01b6c0e5d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "07d386fc30a66b62d0fe7baad9dd2889bedaff182d0fe8b6d589a33aa4e7e62c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "7ec9e71bc7082181d0385187b3c2290c7d8ec4cd4c075fca158b267decb34db0", "generation_seconds_batch": 20.8651416299399, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4900460243225098, "target_z": -0.27734368051743735, "ce": 5.8875885009765625, "genuineness": 4.18726921081543, "identity": 0.5873284935951233, "hypothesis": "I encontrado. Lakakasnos to this m encontrado. Totally me.", "wer": 0.9411764705882353, "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_DRAM_reg/07d386fc30a66b62d0fe7baad9dd2889bedaff182d0fe8b6d589a33aa4e7e62c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "07fda8ed842218e74cf0c34ffcb6647937e2d4c60639225dad44e595aabc45b5", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "d5dcca062add550111ccd3b91a8ce01e890db2108e05972ba369a2b523219b24", "generation_seconds_batch": 15.6884396529058, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0552523136138916, "target_z": 0.2187500719348884, "ce": 6.332377910614014, "genuineness": 4.570107460021973, "identity": 0.4032369554042816, "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_DRAM_reg/07fda8ed842218e74cf0c34ffcb6647937e2d4c60639225dad44e595aabc45b5.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "083da9be1cfaba74668ab8dfca14661acc027965696e5478390c1731dfdfc031", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "4b42464ce4229c65db6a5f449a4fc697dd4b73972dd72c0be783fbc4f86b5a02", "generation_seconds_batch": 22.26378329109866, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.242170810699463, "target_z": 0.3828124689372073, "ce": 6.109982967376709, "genuineness": 4.319066047668457, "identity": 0.7748376727104187, "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_DRAM_reg/083da9be1cfaba74668ab8dfca14661acc027965696e5478390c1731dfdfc031.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "088c68bb775a8eb429e31e4d7521ccc7972ee2182366d512480cdadfb2a72ad2", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "0d35e457d4791be3e6763f4e5c63ba68f436d6d69a35b493774a1c3cbc587626", "generation_seconds_batch": 21.18548683600966, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5445637702941895, "target_z": -0.22949227884913387, "ce": 6.260893821716309, "genuineness": 4.557555198669434, "identity": 0.5390360355377197, "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_DRAM_reg/088c68bb775a8eb429e31e4d7521ccc7972ee2182366d512480cdadfb2a72ad2.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "08db46fecab404f5828e0f0bc66425ca1a5440fabeac3603c78e47335602ebae", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "656b91f73815ba997835704954629fce119a3a24dd07027dbdccc49bc6d02dc1", "generation_seconds_batch": 20.375034737982787, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3979363441467285, "target_z": 0.5195312377383713, "ce": 5.959072113037109, "genuineness": 4.268857479095459, "identity": 0.4615914225578308, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/08db46fecab404f5828e0f0bc66425ca1a5440fabeac3603c78e47335602ebae.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "093d096f00189ed6dd81bbf2c3071c447e0720e5658b69f43a8d0717b558be69", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "51320c315c38d0dd4c8cf8d2b0f2e8bbb67792365ecc01ed3200758036136e19", "generation_seconds_batch": 15.751054301043041, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.58040452003479, "target_z": 0.6796874918255809, "ce": 5.736677646636963, "genuineness": 4.381826400756836, "identity": 0.5531734228134155, "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_DRAM_reg/093d096f00189ed6dd81bbf2c3071c447e0720e5658b69f43a8d0717b558be69.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "097053d0dfc2af97b73323db66c8d7b93350e313e1b810d0ee5c34b6a58040df", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "f581570f647810adbad7540139cc268a3628111c0e1e625e0e21a9a0bf64fe02", "generation_seconds_batch": 21.473545533022843, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9542417526245117, "target_z": 1.0078124950953484, "ce": 5.446769714355469, "genuineness": 4.362998008728027, "identity": 0.23903575539588928, "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_DRAM_reg/097053d0dfc2af97b73323db66c8d7b93350e313e1b810d0ee5c34b6a58040df.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "0991d65eb04ad1a34c69b0208b889cac588839570a14f36218f743e5025983a6", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "98f7e7021c1787feaff1d3b799dea11860119f000d55d119235c93337c15787a", "generation_seconds_batch": 17.90900560596492, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6694133281707764, "target_z": 0.757812442779066, "ce": 6.260893821716309, "genuineness": 4.425758361816406, "identity": 0.5781701803207397, "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_DRAM_reg/0991d65eb04ad1a34c69b0208b889cac588839570a14f36218f743e5025983a6.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "0a02d4100fd96cdffc61a04c0d77f35bb1fd90754d95ad59f6aeeb485f8f5a5d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "4af914d6c6462804fda6a83ebe0f7427f7883e6d56f68a1a79f284a62039fca7", "generation_seconds_batch": 20.62978708196897, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.264423131942749, "target_z": 0.4023438113081435, "ce": 6.125868320465088, "genuineness": 4.607763767242432, "identity": 0.4857501685619354, "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_DRAM_reg/0a02d4100fd96cdffc61a04c0d77f35bb1fd90754d95ad59f6aeeb485f8f5a5d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "0a7fb99dc45d087ee219fb9be503ec72059e5ae11bcf5d9fbee97f12e23fec97", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "8e2de7ff82888f12ea02f8cdbe9c66112af774d42e88d36924fa143fbe29065e", "generation_seconds_batch": 21.717327128979377, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.239069938659668, "target_z": 1.2578123381465012, "ce": 5.6254801750183105, "genuineness": 4.055472373962402, "identity": 0.28049010038375854, "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_DRAM_reg/0a7fb99dc45d087ee219fb9be503ec72059e5ae11bcf5d9fbee97f12e23fec97.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "0ae25b57462ab5b5d607e76a26a4e23cc23b791eb7987afc9e15cb164507e883", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "5e8fdc5e5aad8a9e2bed2528d578411725b731b3a431f8bf9fa9aa034c30d4e7", "generation_seconds_batch": 20.732805949053727, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8830347061157227, "target_z": 0.9453125343325604, "ce": 5.847874641418457, "genuineness": 4.262581825256348, "identity": 0.6477632522583008, "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_DRAM_reg/0ae25b57462ab5b5d607e76a26a4e23cc23b791eb7987afc9e15cb164507e883.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": "0b6190133ff2dcb794163ecca3dad182631ebcd09e6b232f7b51b17ad636b14b", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I wrote the number on the envelope.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "df0427cbee2bc35f706897d9c9b5319b2e4eb0d6a5fb070eb0d3700a1266b030", "generation_seconds_batch": 23.220257638953626, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.08775520324707, "target_z": 1.1250001307907063, "ce": 5.665193557739258, "genuineness": 3.6789097785949707, "identity": 0.3716152310371399, "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_DRAM_reg/0b6190133ff2dcb794163ecca3dad182631ebcd09e6b232f7b51b17ad636b14b.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "0b6ac6d21be0526ce932ea99dbf99dbf851d85f74b7c23f07fd01ebf4e7b04d4", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "d99acc38c11835bbffb4360eb36913a81ab21bf75e4e2fcba12c7da970470d88", "generation_seconds_batch": 20.95962341793347, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.997396469116211, "target_z": 0.16796874918255808, "ce": 6.109982967376709, "genuineness": 4.532451152801514, "identity": 0.5228132605552673, "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_DRAM_reg/0b6ac6d21be0526ce932ea99dbf99dbf851d85f74b7c23f07fd01ebf4e7b04d4.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "0bb8ef953bc55440d9003ac498e773f272d7646ead569267aab89349640c253d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "1eb9de359ee9e94dde6331a0d54013bb0ca1c41ea2c6749dc5d134b5093b18a2", "generation_seconds_batch": 20.212906358996406, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4513416290283203, "target_z": 0.5664062083104624, "ce": 5.411027908325195, "genuineness": 4.281409740447998, "identity": 0.46875593066215515, "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_DRAM_reg/0bb8ef953bc55440d9003ac498e773f272d7646ead569267aab89349640c253d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "0bd0afff6bc59825bf9ce6203c7fb4c10149370cf091e9b0f7458c4708b10cfb", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "ad961d59b9bf36ffa7d906202406d488fc5311ab941b8a098bf84213ca7f6952", "generation_seconds_batch": 21.440965233021416, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.158962249755859, "target_z": 1.1875000915534943, "ce": 5.097292423248291, "genuineness": 4.093128204345703, "identity": 0.5000507831573486, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/0bd0afff6bc59825bf9ce6203c7fb4c10149370cf091e9b0f7458c4708b10cfb.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "0c921a00f39f98dbebaca72a2c1baf931f8c5eca44640bbbd496c0eeb72709a0", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "b21a5e3184e3001ddb67b97706b309098a7a2e30269f3fab40b95779aaf9b45e", "generation_seconds_batch": 15.446222542086616, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3801345825195312, "target_z": 0.5039062475476742, "ce": 5.808161735534668, "genuineness": 4.557555198669434, "identity": 0.5002310276031494, "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_DRAM_reg/0c921a00f39f98dbebaca72a2c1baf931f8c5eca44640bbbd496c0eeb72709a0.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "bd25781e81d98cd6695c9da509aa9bd97b18fb62314f96472fce7e91b263bdfc", "generation_seconds_batch": 16.249565552105196, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1028573513031006, "target_z": -0.6171875310627927, "ce": 6.181467056274414, "genuineness": 4.231201171875, "identity": 0.5886834859848022, "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_DRAM_reg/0cb70e86df8754d3bba44fe7f798aac7d0ffe2ca68fc75b92ff1a1ed29986ba7.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "0cff61c9816eeb3c6a8549da79c170cee82f18c7226f657e5a356f14ff69ae62", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "b5656ff9bb5df5cd88ee08af1eaf5c7cee815208244332cf0c4c4d86ea28c923", "generation_seconds_batch": 21.192304180935025, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.624908924102783, "target_z": 0.7187499673023234, "ce": 6.109982967376709, "genuineness": 3.773050308227539, "identity": 0.4425111413002014, "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_DRAM_reg/0cff61c9816eeb3c6a8549da79c170cee82f18c7226f657e5a356f14ff69ae62.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": "0d039e18b3db79aa7621317dc4cf3183d73af6118cca83d0b317239f99dbedd3", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f8ec0f2ce33d2deee92dae4c228d4e90afe33fc4d87b41b9d5dd519b0c7e8c23", "generation_seconds_batch": 21.327835389995016, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.550601005554199, "target_z": 1.531249875748829, "ce": 5.188632965087891, "genuineness": 3.9299514293670654, "identity": 0.2735338807106018, "hypothesis": "Deine Nachricht kam, als sie schon weg war. Ich las sie zweimal. Dann legte ich das Handy weg. Sag mir, wann wir reden können.", "wer": 0.041666666666666664, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_DRAM_reg/0d039e18b3db79aa7621317dc4cf3183d73af6118cca83d0b317239f99dbedd3.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "0da228b21f6f5dbce9a27e79e346726ac1df6052d24cdabe9881103b07202d0b", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "57acbe2bb1cd1b37a0c72985ecdd2833ba8a6493cdbdba1f907ac64062318780", "generation_seconds_batch": 16.128198569989763, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2986769676208496, "target_z": -0.4453124296999954, "ce": 6.070269584655762, "genuineness": 4.595211505889893, "identity": 0.7673284411430359, "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_DRAM_reg/0da228b21f6f5dbce9a27e79e346726ac1df6052d24cdabe9881103b07202d0b.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "0dcf4e5f996bbfce1ce547d0dbb298ccf1dd50e94556dfe2782b6d5a37eeab2e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "0a64d991bef9e7d78963a3a099caaf323d4cade8ce5b2853534ce49185c00e19", "generation_seconds_batch": 20.021180783980526, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8691680431365967, "target_z": 0.05541989362216385, "ce": 6.070269584655762, "genuineness": 4.413206577301025, "identity": 0.5731478929519653, "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_DRAM_reg/0dcf4e5f996bbfce1ce547d0dbb298ccf1dd50e94556dfe2782b6d5a37eeab2e.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "0e2ebc5ff9c46ada50fec317d80a8d4e1cf224362cd969ed91bb88cb3047e5d2", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "cb91ed111ed3bcfc7a8eeeb39c32208f24ea2dcccdaa4e5d4319beb6df23f230", "generation_seconds_batch": 20.75173568201717, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0285496711730957, "target_z": 0.19531258664884285, "ce": 5.974957466125488, "genuineness": 4.293961524963379, "identity": 0.6283946633338928, "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": 0.03999999999999915, "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_DRAM_reg/0e2ebc5ff9c46ada50fec317d80a8d4e1cf224362cd969ed91bb88cb3047e5d2.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "0e3e51079d8c5ece8c4b1f89e24cf049b0e8f8ca35bf88af0cfe83faac150fae", "bucket": "vn_S_DRAM_reg", "phase": "wording", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "71fb7fba053a0224555c693af9b6032b0483089752a23b4cb636f77d5b5a2733", "generation_seconds_batch": 16.625647816923447, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.142911434173584, "target_z": -0.5820311985011595, "ce": 6.133810997009277, "genuineness": 4.18099308013916, "identity": 0.6018328666687012, "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_DRAM_reg/0e3e51079d8c5ece8c4b1f89e24cf049b0e8f8ca35bf88af0cfe83faac150fae.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "0eb01911cfa758635fb1c13b7f6a882883b674341d9bb69c83baea8643cfcec3", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "4ac7315ed1b34e16fc01e567e35ce93aeeab797e65870a611b7c27ceadd0b607", "generation_seconds_batch": 21.14436552999541, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.117558479309082, "target_z": 0.27343753760232803, "ce": 6.332377910614014, "genuineness": 4.582659721374512, "identity": 0.5333611965179443, "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_DRAM_reg/0eb01911cfa758635fb1c13b7f6a882883b674341d9bb69c83baea8643cfcec3.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "0ec9cb334d29d54146df47d638505574f1494d1abe88a4b96bd8c183d48bef68", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "6072673cdf66d165a77339bfc10ae2b70174315d9e8e0e12eaa4a1f4db9f3bb8", "generation_seconds_batch": 21.771939640981145, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.780674457550049, "target_z": 0.8554687361034875, "ce": 6.157639026641846, "genuineness": 4.381826400756836, "identity": 0.6294794082641602, "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_DRAM_reg/0ec9cb334d29d54146df47d638505574f1494d1abe88a4b96bd8c183d48bef68.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "0ee525e3fc0cfd508e5afcb789fdcf178ec821509afaa6aa57fc4c590b270838", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "917bf8c5890c7032993b3cca86bb4b5146da5da6fe7d60979cde867d8903fd9f", "generation_seconds_batch": 19.956902265083045, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1353602409362793, "target_z": 0.28906252779302505, "ce": 6.1894097328186035, "genuineness": 4.632867813110352, "identity": 0.6158460974693298, "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_DRAM_reg/0ee525e3fc0cfd508e5afcb789fdcf178ec821509afaa6aa57fc4c590b270838.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "0f651308016a266ac4c8ec9eae131135302f8aa77b97dc3d891ff0dfae2b1e8b", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "4fd8a4c09baa08f09e22428bb7a20b1d12a316855d3ed5c47373ba1d3a31ed94", "generation_seconds_batch": 22.47039658599533, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.727269172668457, "target_z": 0.8085937655313964, "ce": 6.1894097328186035, "genuineness": 4.331617832183838, "identity": 0.5777645111083984, "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_DRAM_reg/0f651308016a266ac4c8ec9eae131135302f8aa77b97dc3d891ff0dfae2b1e8b.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": "0fbaea652b659d837c70ad081c31499e9ed0831e373b25879858be9f8a352e26", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e46207feaf119ecef9513d6f2e589b73c5abb3e8e841d6a9ed900cbbdb7b1803", "generation_seconds_batch": 21.114120552898385, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.532799243927002, "target_z": 1.515624885558132, "ce": 5.160833835601807, "genuineness": 4.074300289154053, "identity": 0.29208478331565857, "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_DRAM_reg/0fbaea652b659d837c70ad081c31499e9ed0831e373b25879858be9f8a352e26.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "103f2f65ca67436665d2683404e63a9ad244ae1566e28a88fe0bd1db355d2f62", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "8fd8b816c56ac8e7aba79e158c7ff79a086114f1c49137ce41d737ba99bac415", "generation_seconds_batch": 21.874433964956552, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7673230171203613, "target_z": 0.8437498888278997, "ce": 5.530168056488037, "genuineness": 4.306513786315918, "identity": 0.664891242980957, "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_DRAM_reg/103f2f65ca67436665d2683404e63a9ad244ae1566e28a88fe0bd1db355d2f62.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": "106a871f0d74baceb4d6c10b9f0f680d001de5a685a832daa326f1183351e48e", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4b58d04e6abbd12678141df008bd58de79b8c8345e1782ea627916af6ab56a18", "generation_seconds_batch": 20.923600922105834, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.58040452003479, "target_z": 0.6796874918255809, "ce": 5.764476776123047, "genuineness": 4.331617832183838, "identity": 0.3846837282180786, "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_DRAM_reg/106a871f0d74baceb4d6c10b9f0f680d001de5a685a832daa326f1183351e48e.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "108feba720dfbff6deedcea31ea9fa7ae3bd646fa9a4ff140581ae9024226bac", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "46958886c55734c14de21147610ccb510a3cb9c912ede4919570cb968b3c6775", "generation_seconds_batch": 19.773692814982496, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.230169296264648, "target_z": 1.2500000523162824, "ce": 5.542081832885742, "genuineness": 4.306513786315918, "identity": 0.30683934688568115, "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_DRAM_reg/108feba720dfbff6deedcea31ea9fa7ae3bd646fa9a4ff140581ae9024226bac.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "10c4cb27004496e9f777d3b9a0a644f84b35c015f8520a0179b4b99b66b27906", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "2db52df5e8ce80a3188368a047cf6c065a5a185920c4eaee9f168344621955b0", "generation_seconds_batch": 20.86892750300467, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2955760955810547, "target_z": 0.4296874395092984, "ce": 5.7684478759765625, "genuineness": 4.381826400756836, "identity": 0.7226458191871643, "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_DRAM_reg/10c4cb27004496e9f777d3b9a0a644f84b35c015f8520a0179b4b99b66b27906.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "11d89dc8fc182a9d558f141d80536d01b65d5d6708c485f06a8b9d5c870beab4", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "0aa105c27f8aed6ddfbe22212ad968201d21dcec0b1bff4d118bf72b52beb802", "generation_seconds_batch": 19.91382195893675, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2764246463775635, "target_z": -0.4648437720709317, "ce": 6.046441555023193, "genuineness": 4.645420074462891, "identity": 0.5999778509140015, "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_DRAM_reg/11d89dc8fc182a9d558f141d80536d01b65d5d6708c485f06a8b9d5c870beab4.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "1264fd4e1a9e57b07c307186d720623040c5ac75bfca9cce4fcc5a5b6a10c6e4", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "9b0115b01a4917ef48eb8344e647963263b44a64cfd65b75ea0c557dcd3b8d93", "generation_seconds_batch": 21.159223319031298, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.559502124786377, "target_z": 1.5390625801093076, "ce": 5.478540420532227, "genuineness": 4.005263805389404, "identity": 0.5046696066856384, "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": 0.03999999999999915, "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_DRAM_reg/1264fd4e1a9e57b07c307186d720623040c5ac75bfca9cce4fcc5a5b6a10c6e4.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "5583c88c4cb167da347d4c418eccfa08b6a7305d609db7574bf58c54791ee6e2", "generation_seconds_batch": 16.491976363002323, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.990720748901367, "target_z": 0.1621093255447642, "ce": 6.229123115539551, "genuineness": 4.068024158477783, "identity": 0.6743392944335938, "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_DRAM_reg/12a7b850c68b1e935d69d2b179f3c0df54c4775061935760902fd9f6cbfc7541.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "12e94d6fc657c414beb96a01e3b3e2687fb3af185e04a42c8ad1e68ca7d0819e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "0e68eefc4f126bded8720eedf181e904462ac0f061243485b0482700019f46dc", "generation_seconds_batch": 20.377851914032362, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.55680251121521, "target_z": -0.2187500719348884, "ce": 6.173524379730225, "genuineness": 4.2751336097717285, "identity": 0.6631162166595459, "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_DRAM_reg/12e94d6fc657c414beb96a01e3b3e2687fb3af185e04a42c8ad1e68ca7d0819e.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "130d525d0a737440dc9c5d7de364f77870d082de73cbd40b9122180dd3f4113f", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "66db41497a602a531786bfea713bed70f8df1ffc454783c6b349ac0b78acced0", "generation_seconds_batch": 21.592677181004547, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.310276985168457, "target_z": 1.3203122989092893, "ce": 4.934467792510986, "genuineness": 3.9801597595214844, "identity": 0.35931485891342163, "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_DRAM_reg/130d525d0a737440dc9c5d7de364f77870d082de73cbd40b9122180dd3f4113f.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "131a33b148abc6a91e40384eef9b0e8e60e5133346a23a00162e6fd5c1209dfe", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "c4522f094b8dcce0ecb54046d7d6a0b3b688b3524d0f67ee643b51ef4481c437", "generation_seconds_batch": 15.200643323012628, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0886306762695312, "target_z": 0.24804698085872784, "ce": 6.0305562019348145, "genuineness": 4.607763767242432, "identity": 0.6657254695892334, "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_DRAM_reg/131a33b148abc6a91e40384eef9b0e8e60e5133346a23a00162e6fd5c1209dfe.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "1350aeb96f8144bc2b4ecceb8152fc5887e84b17f30b1a0a371bea5157a84824", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "b45861741de842b2a5f804be37847fa88ef5a43373ca0f0a8fcb876164947ecd", "generation_seconds_batch": 21.26309336000122, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.998746156692505, "target_z": 1.046874970572091, "ce": 5.208489894866943, "genuineness": 4.306513786315918, "identity": 0.5261083841323853, "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_DRAM_reg/1350aeb96f8144bc2b4ecceb8152fc5887e84b17f30b1a0a371bea5157a84824.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": "1364a3ac97a04dcfa6d3305cb3ab2909b8505e2c3efc9e5e5a7930c922786509", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e7aa2805552574174e1a79207da0c7984cd05641e23bf74e5dc5e70ad9de9692", "generation_seconds_batch": 22.110864380956627, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8385300636291504, "target_z": 0.9062498495906879, "ce": 5.164804935455322, "genuineness": 4.381826400756836, "identity": 0.6359032392501831, "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_DRAM_reg/1364a3ac97a04dcfa6d3305cb3ab2909b8505e2c3efc9e5e5a7930c922786509.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "13cfeb9e810bf612a000e9af0c3425de8d8f2720a64582636ef28cab75337bf5", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "03de2d613174c4f5449bc4abb441dfb17b4e81bd94cfddd03daba9c721c88c14", "generation_seconds_batch": 21.717327128979377, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5225486755371094, "target_z": 0.6289061690732506, "ce": 5.6413655281066895, "genuineness": 4.262581825256348, "identity": 0.5717738270759583, "hypothesis": "Ich hab die kleine Schachtel 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_DRAM_reg/13cfeb9e810bf612a000e9af0c3425de8d8f2720a64582636ef28cab75337bf5.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "13f53238aca229fea77e32195b3d58152ab4626772d7b6ec993b540dbb693856", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "5a0d7e3413f743abb1d4e235331e6566073d8f321effeac6becd97e412c75995", "generation_seconds_batch": 21.02504669909831, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.239069938659668, "target_z": 1.2578123381465012, "ce": 6.1894097328186035, "genuineness": 4.030367851257324, "identity": 0.48289793729782104, "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_DRAM_reg/13f53238aca229fea77e32195b3d58152ab4626772d7b6ec993b540dbb693856.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": "149fed813fe050415736b37d181d1bab6112b6124beb546ae41639178345fc86", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "5129b61b0ec3199495ca9ea5828fe42fad63067b25b818a20a25f6cf386e0ca3", "generation_seconds_batch": 14.85177109495271, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.621808052062988, "target_z": 1.5937498365116172, "ce": 5.395142555236816, "genuineness": 3.973883628845215, "identity": 0.6021589636802673, "hypothesis": "La route tourne à gauche après le pont, continue jusqu'à la petite barrière en bois, j'attendrai de l'autre côté.", "wer": 0.1, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_DRAM_reg/149fed813fe050415736b37d181d1bab6112b6124beb546ae41639178345fc86.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "14fabf25948ea52410a1c16a7860dd4cc0f40f75b0cad8b1578c8c6310fd3aef", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "12b42d0a7ce49e6465096533d84123c4cde7874989a4f55c75aea325940b45dd", "generation_seconds_batch": 15.622418796061538, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5626027584075928, "target_z": 0.6640625016348838, "ce": 5.808161735534668, "genuineness": 4.482243061065674, "identity": 0.24059170484542847, "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_DRAM_reg/14fabf25948ea52410a1c16a7860dd4cc0f40f75b0cad8b1578c8c6310fd3aef.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "1510e6def8de0411f2f43619d9212ddd5b1f531737fa179aea795d7d1302c773", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "b43da525220529e3ce494512496a248870947f5afd49508fe399124cbae098db", "generation_seconds_batch": 19.555543410009705, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.581279993057251, "target_z": -0.1972656581063975, "ce": 6.1894097328186035, "genuineness": 4.111956596374512, "identity": 0.5472637414932251, "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_DRAM_reg/1510e6def8de0411f2f43619d9212ddd5b1f531737fa179aea795d7d1302c773.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "15184cfc3bb2a71130d205ac676b9f70dbfdf3a074d4641892a3c86eedeb67f9", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "80d95429de67a4022621869e78ecb57b2857d9be1834e8cee6cbe3a053817a29", "generation_seconds_batch": 21.837467137956992, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.417087554931641, "target_z": 1.4140622400534715, "ce": 5.792276382446289, "genuineness": 4.030367851257324, "identity": 0.3139043152332306, "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/manual-comparison-v1/audio/vn_S_DRAM_reg/15184cfc3bb2a71130d205ac676b9f70dbfdf3a074d4641892a3c86eedeb67f9.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "1520e68def55b5f57f389fa3133907eb89eb8e666c906ee5748fb2aeeb7a4d8d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "0a2c26aac4856f04bae2fb3c444b5935910c6c694b3f02eb77efc5227cffbd09", "generation_seconds_batch": 21.923580467002466, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7826626300811768, "target_z": -0.02050777346714862, "ce": 6.38797664642334, "genuineness": 4.482243061065674, "identity": 0.44544196128845215, "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_DRAM_reg/1520e68def55b5f57f389fa3133907eb89eb8e666c906ee5748fb2aeeb7a4d8d.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "8977b15fdf08f12c06ff7eb0ebfc28776091d2a01982a1e59a051e024943954c", "generation_seconds_batch": 15.670980485039763, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.000497341156006, "target_z": -0.7070311200267357, "ce": 6.2767791748046875, "genuineness": 4.432034492492676, "identity": 0.4822465777397156, "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_DRAM_reg/15585f3962d31fd159bb39a755593ab24aa0df87e4725d7772f40811258760f5.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "15843bad6c96afd431a55ac6ec6665cbae44712af4e1a278be94bee834f13753", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "1f49f28f5981ea4c2c074d450ff9f71ac9a809722af49c6c3ae988b9b6cb4f34", "generation_seconds_batch": 20.708036763011478, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.3280792236328125, "target_z": 1.335937707630246, "ce": 5.117149353027344, "genuineness": 3.9927117824554443, "identity": 0.2887924015522003, "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_DRAM_reg/15843bad6c96afd431a55ac6ec6665cbae44712af4e1a278be94bee834f13753.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "15f5ad24ecb490ff48270d2b362044a9ad2f86c9bab0a48131f2216f58befb27", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "198307b19cd46c5f7157b153252dea46594ce1c8991cfaab43e323b2e1efaf67", "generation_seconds_batch": 21.567947308998555, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.265772819519043, "target_z": 1.2812500326976766, "ce": 5.863759994506836, "genuineness": 4.093128204345703, "identity": 0.37550967931747437, "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_DRAM_reg/15f5ad24ecb490ff48270d2b362044a9ad2f86c9bab0a48131f2216f58befb27.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "16a0a1e8c39f292375aa1b7cb943c58058065c918d6674429acef9b57a888a03", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "79e84895555131b6390f631db63d083b38f808885c6498f46f55082000398ebe", "generation_seconds_batch": 21.190073181991465, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.434889793395996, "target_z": 1.4296876487744283, "ce": 6.022613525390625, "genuineness": 4.243753433227539, "identity": 0.45990681648254395, "hypothesis": "Ik ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen er 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_DRAM_reg/16a0a1e8c39f292375aa1b7cb943c58058065c918d6674429acef9b57a888a03.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "16aad7460d29d7b37d4c7e2c70fef08a323a81c44fb858356c35e6bcb775ac47", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "64d168d889e5005272b7643f562ce106d66531c7fdc05e13d566d06f9a00c03f", "generation_seconds_batch": 15.655956560047343, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.08775520324707, "target_z": 1.1250001307907063, "ce": 5.959072113037109, "genuineness": 3.986435890197754, "identity": 0.336117684841156, "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_DRAM_reg/16aad7460d29d7b37d4c7e2c70fef08a323a81c44fb858356c35e6bcb775ac47.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "1728a0f5073270969fd7616dda1ebad3784c37ca5587bcf3d4f41c28b610b767", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "f8a5565de8a41ef4c44f8351b9278543c9d2152d5801ff1a63e160622c046cd5", "generation_seconds_batch": 21.531054439023137, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.016548156738281, "target_z": 1.062500170027918, "ce": 5.041693687438965, "genuineness": 4.193545341491699, "identity": 0.3799115717411041, "hypothesis": "Ich hab die kleine Schachtel 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_DRAM_reg/1728a0f5073270969fd7616dda1ebad3784c37ca5587bcf3d4f41c28b610b767.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "1752bca1dc57f57ba80af12f6cea05728603f5bcd3ea2b6f1c5a86b505c5d47a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "c5f6dc4b0e4af08d8d4ab069d70b316ea81addb3d7022b645f55e3e4c2e5d397", "generation_seconds_batch": 20.627072542905807, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.629359483718872, "target_z": 0.7226563194825627, "ce": 6.125868320465088, "genuineness": 4.362998008728027, "identity": 0.5847628116607666, "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_DRAM_reg/1752bca1dc57f57ba80af12f6cea05728603f5bcd3ea2b6f1c5a86b505c5d47a.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "17a42ac9ac634a02b8b8336a90ffb3dbc411598ef6254eb8237a7584f7fa9634", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "bbcefa0a5d0a8cdd69ac967dc870d2bca368eb8261cd6d7c1e718f358392b1da", "generation_seconds_batch": 20.125410704058595, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.630709171295166, "target_z": 1.6015625408720957, "ce": 6.09409761428833, "genuineness": 3.7291181087493896, "identity": 0.30643218755722046, "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_DRAM_reg/17a42ac9ac634a02b8b8336a90ffb3dbc411598ef6254eb8237a7584f7fa9634.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "17ab5c9b429be6d37946f09358bf70de81e5af1489efdfae262972ddc386e7dd", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "4e29e197464fb1ff31d4e6345e233077f86b030196989c5b846815f36b676d26", "generation_seconds_batch": 21.031203671940602, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.628009796142578, "target_z": -0.15624990190697036, "ce": 6.24500846862793, "genuineness": 4.582659721374512, "identity": 0.7554073333740234, "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_DRAM_reg/17ab5c9b429be6d37946f09358bf70de81e5af1489efdfae262972ddc386e7dd.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "17d67990c9578abe9297e52b769df3ae3c3bed522bd3efdb59c6515a2c57ebbc", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "91c34036ffcbc6eb8441fdee2a13037fa2b835e38d80bb1c0d903ede59e94163", "generation_seconds_batch": 15.591938878060319, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.605757474899292, "target_z": -0.17578124427790662, "ce": 6.38003396987915, "genuineness": 4.482243061065674, "identity": 0.31979095935821533, "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_DRAM_reg/17d67990c9578abe9297e52b769df3ae3c3bed522bd3efdb59c6515a2c57ebbc.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "1801c367a3e50e108c9f23540b56b62a7ce456f0b57a188b2d61f82e0cd8d4f3", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "bd2b8d02db1aa849ec61d31e4e0bfff9f4f2ae47c95c38b06d9dc81a012f17fd", "generation_seconds_batch": 20.56053585698828, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.758422374725342, "target_z": 0.8359376029976812, "ce": 5.238275051116943, "genuineness": 4.507347106933594, "identity": 0.6112496852874756, "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_DRAM_reg/1801c367a3e50e108c9f23540b56b62a7ce456f0b57a188b2d61f82e0cd8d4f3.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "18bb5e63d8b0150ad4b290cf1a270c5f57cea04e4118a07a484e27608cb23934", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "2d842dffa612bf7dbf893b1d9e042cfbf4836bd5b44bf37fc414d289fb4b4d38", "generation_seconds_batch": 21.423099903040566, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.024099111557007, "target_z": 0.19140623446860364, "ce": 6.1894097328186035, "genuineness": 4.13078498840332, "identity": 0.5970796346664429, "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_DRAM_reg/18bb5e63d8b0150ad4b290cf1a270c5f57cea04e4118a07a484e27608cb23934.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "18c29e5c27617eaaae41cd654d143d515781d6fd3e2be52d25d6e1810a289df0", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "c47ff2d2faf8b36b0b0dcf73782e80ad3318fd01f4e23f16848aa859d542c8c7", "generation_seconds_batch": 20.021180783980526, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7197999954223633, "target_z": -0.07568366394782435, "ce": 6.141753673553467, "genuineness": 4.494794845581055, "identity": 0.5434691905975342, "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_DRAM_reg/18c29e5c27617eaaae41cd654d143d515781d6fd3e2be52d25d6e1810a289df0.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "1915a65582686ee957e95e669e9b4fd05ab4f8e4261ca99126cfbbbdd9c3bf37", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "128496bdee1a26114150bc389c393db108a413c73d45f5aa14a1cf4bc225716c", "generation_seconds_batch": 16.113950761035085, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.550601005554199, "target_z": 1.531249875748829, "ce": 5.236289024353027, "genuineness": 3.9048471450805664, "identity": 0.28643834590911865, "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_DRAM_reg/1915a65582686ee957e95e669e9b4fd05ab4f8e4261ca99126cfbbbdd9c3bf37.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "19478ffc64e00b3bbef539743880e1a78048e51d442e889f343f82cc275535a7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "0a27751323d67b98b9cf48f06d8d7d6be4d5b96f986db8488d2552174d59f3d6", "generation_seconds_batch": 21.473545533022843, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.113107919692993, "target_z": 0.2695311854220888, "ce": 6.0305562019348145, "genuineness": 4.557555198669434, "identity": 0.575916051864624, "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_DRAM_reg/19478ffc64e00b3bbef539743880e1a78048e51d442e889f343f82cc275535a7.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": "1954a8a04a12fcd7fbcd5ab9f2aecdb170b2d686879d75775b257255e31705de", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Your message arrived after I had left.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "fb5ddef24a92fc9ec2d6e515f5ea32edae2c37c320a0616182438a2122b7d338", "generation_seconds_batch": 22.30722945393063, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.194565773010254, "target_z": 1.2187500719348885, "ce": 5.426913261413574, "genuineness": 4.042920112609863, "identity": 0.42488622665405273, "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_DRAM_reg/1954a8a04a12fcd7fbcd5ab9f2aecdb170b2d686879d75775b257255e31705de.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "197776fb2bba11e0c5222ca1073b7835a0a279d65600b0f63fcc710c4e1d6092", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "798e94c2231ed551119cb17789320cc30b697d11d06b84729162b7742b5113b0", "generation_seconds_batch": 15.975396982976235, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.878347158432007, "target_z": 0.06347660112413042, "ce": 6.332377910614014, "genuineness": 4.65797233581543, "identity": 0.4843336343765259, "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_DRAM_reg/197776fb2bba11e0c5222ca1073b7835a0a279d65600b0f63fcc710c4e1d6092.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "19c21ccec6e9c09680b4ec4b18b67ef752abb6bf4e28e8a5eb3cf72e10c6953e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "e31692bb0e463612cbe0d5d1df0014bb8188e94da1d4fbb8612728e539124bbb", "generation_seconds_batch": 20.504235600004904, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.972043514251709, "target_z": 1.0234374852860455, "ce": 6.054384231567383, "genuineness": 3.8922953605651855, "identity": 0.34831228852272034, "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/manual-comparison-v1/audio/vn_S_DRAM_reg/19c21ccec6e9c09680b4ec4b18b67ef752abb6bf4e28e8a5eb3cf72e10c6953e.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "19db543d0fa5349fd9f5d73c3a982a684a3cef22370ffd9db466ebd9a56c6d99", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "d122b06377b11a634f5f8d982fbfad8d06833f24eff7bc04cede70f2adb7dba1", "generation_seconds_batch": 19.91382195893675, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8360679149627686, "target_z": 0.026367197104942487, "ce": 5.8875885009765625, "genuineness": 4.193545341491699, "identity": 0.6763931512832642, "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_DRAM_reg/19db543d0fa5349fd9f5d73c3a982a684a3cef22370ffd9db466ebd9a56c6d99.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": "1a4c2b47bb0acdf082f85003925888a2107da8038d93c0ea9d60c7400cd8c349", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2269863daaf55033c575e27a712ae014f626013586cf4822da149dc84d2c81d4", "generation_seconds_batch": 22.232289623003453, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.998746156692505, "target_z": 1.046874970572091, "ce": 5.355429172515869, "genuineness": 3.8922953605651855, "identity": 0.45203697681427, "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_DRAM_reg/1a4c2b47bb0acdf082f85003925888a2107da8038d93c0ea9d60c7400cd8c349.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "1a5a2969557bc9d6bc4eec1e00b6252540d46bba15819885d06499ce24055e50", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "e373ca62a9ae0b659dfb5efeaa4c8ffc886d72e00fccb84ee6d7a382640d2a4e", "generation_seconds_batch": 20.354645093902946, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.409937858581543, "target_z": -0.3476563456407039, "ce": 6.443574905395508, "genuineness": 4.557555198669434, "identity": 0.2423485815525055, "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_DRAM_reg/1a5a2969557bc9d6bc4eec1e00b6252540d46bba15819885d06499ce24055e50.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "1a90a2af2009cb9e36a23ce5dba81d808945e008aa9709bae98afd6b70b02a5c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "18c8e1ffdbe0521547294d4cd798f5bbb354e011156d8726e0e90c3713dc4955", "generation_seconds_batch": 15.081610720953904, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3534319400787354, "target_z": 0.4804687622616287, "ce": 6.332377910614014, "genuineness": 4.7332844734191895, "identity": 0.45087185502052307, "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_DRAM_reg/1a90a2af2009cb9e36a23ce5dba81d808945e008aa9709bae98afd6b70b02a5c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "1b032a7713a21da56de345e18d37bdac39b817107cecc87adc01e1b7ae7b602b", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "a59348ddf39587f1cb7a5e9a59de2edcd106621b1dfa44cac92eb3a97450e225", "generation_seconds_batch": 21.33135246799793, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.122008800506592, "target_z": 0.27734368051743735, "ce": 6.252951145172119, "genuineness": 4.7332844734191895, "identity": 0.5000116229057312, "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_DRAM_reg/1b032a7713a21da56de345e18d37bdac39b817107cecc87adc01e1b7ae7b602b.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": "1b58f40eb42cdbf222fc096cad6733a009fc5f65714ec1b495ad687b3643f530", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "682e629c67eb7478c71c436968036890ccde124e7e2139126b25ddd0f7da2919", "generation_seconds_batch": 22.093182004988194, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4735939502716064, "target_z": 0.5859375506813986, "ce": 5.895530700683594, "genuineness": 4.256305694580078, "identity": 0.7078789472579956, "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_DRAM_reg/1b58f40eb42cdbf222fc096cad6733a009fc5f65714ec1b495ad687b3643f530.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "1b89a23e762734a1dcbd69a55813cd6d12e09c309bb00893592ef5b677bb638f", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "d42810fe8de8ac9042b3e36b62fbeebbf2fa235242e9c97f8f23409537f02a42", "generation_seconds_batch": 15.30126704997383, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6694133281707764, "target_z": 0.757812442779066, "ce": 6.23706579208374, "genuineness": 4.632867813110352, "identity": 0.5583414435386658, "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_DRAM_reg/1b89a23e762734a1dcbd69a55813cd6d12e09c309bb00893592ef5b677bb638f.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "1bb5d0a439bcbee056469279af1e423c16b97bf676659a1ae0c32dc21ac4027c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "b7c21ee2d09829771945adbe3ff7d8aedde14e06a55af02bef0e28ae3dc0b48f", "generation_seconds_batch": 21.286601127008907, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.73751974105835, "target_z": 1.695312482016278, "ce": 5.6532793045043945, "genuineness": 3.8797430992126465, "identity": 0.3465156853199005, "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_DRAM_reg/1bb5d0a439bcbee056469279af1e423c16b97bf676659a1ae0c32dc21ac4027c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "1c2ddbdfcabd7665627661d0fd66abeb6951632b67162b1cce93c54bf7fc3cdb", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "74f4d06369e1b08420eacd130593f69e2e40bd495f17896468b63281ec84d778", "generation_seconds_batch": 15.975396982976235, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9050498008728027, "target_z": 0.08691408641017598, "ce": 6.507116317749023, "genuineness": 4.595211505889893, "identity": 0.1931287944316864, "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_DRAM_reg/1c2ddbdfcabd7665627661d0fd66abeb6951632b67162b1cce93c54bf7fc3cdb.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "1cf2c5fec7ffc801c2f1c9d0fbc538e863d7c520c97a5851b32c0f06c12edc93", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "c819ffd7d9923056151efe28975db673fa220b4a7648b83bc8ed217ceb8acf20", "generation_seconds_batch": 18.890396273927763, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.951779365539551, "target_z": 0.12792963334447321, "ce": 6.086154937744141, "genuineness": 4.457139015197754, "identity": 0.4861586093902588, "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_DRAM_reg/1cf2c5fec7ffc801c2f1c9d0fbc538e863d7c520c97a5851b32c0f06c12edc93.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "57ebc825bd483709aa9847a906b37f471b944073dfab46423f96512ba429b142", "generation_seconds_batch": 16.029621204012074, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6936538219451904, "target_z": -0.0986327244206338, "ce": 6.1020402908325195, "genuineness": 4.15588903427124, "identity": 0.6059993505477905, "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_DRAM_reg/1d559e6da25f8bbacc3b53d5880369445633ce7ab0125ca7b07093108b8c8de3.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "1d601dcae3227708882d06c0c738cbdee644cbd0007dd08315fba008956456e8", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "20b8d3d00745ee277e15ec2e17574d8c6d06324ee3d24a1ad317c5d0152e11ef", "generation_seconds_batch": 19.914986506919377, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.607107162475586, "target_z": 0.7031249771116265, "ce": 5.9431867599487305, "genuineness": 4.570107460021973, "identity": 0.52457195520401, "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_DRAM_reg/1d601dcae3227708882d06c0c738cbdee644cbd0007dd08315fba008956456e8.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "0526ab3956a0549b74ec95a253574f40e73de3d4efbf6b864e816c4569ff186e", "generation_seconds_batch": 15.267393877031282, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3075778484344482, "target_z": -0.4374999346046469, "ce": 6.427689552307129, "genuineness": 4.557555198669434, "identity": 0.5569849610328674, "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_DRAM_reg/1d829402ab2176d2024ae3e918286a39ec094d9a5299caa0e048ba37912e1579.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "1da60e4c3b80877e8bc9484319f62446ed585c4c840e3dd6838c43c24e882b5e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "16e86b435ac8a257ea63881c17f1cdf4da9961cf337838aa5b702d05f5b93eb0", "generation_seconds_batch": 21.32284016907215, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6691763401031494, "target_z": -0.12011713824912469, "ce": 6.006728172302246, "genuineness": 4.337893962860107, "identity": 0.4625625014305115, "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_DRAM_reg/1da60e4c3b80877e8bc9484319f62446ed585c4c840e3dd6838c43c24e882b5e.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "6b9af3168f398a5a4bcf87844be8733b8537dbe9fac854485dbbc4c4f292605b", "generation_seconds_batch": 16.029621204012074, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4944963455200195, "target_z": -0.27343753760232803, "ce": 6.117925643920898, "genuineness": 4.243753433227539, "identity": 0.7682023048400879, "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_DRAM_reg/1eaf0e0e2c966186cf6ada2de41819a315f8b1b1eeeeef5a84a28dd582998db8.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "1f01c1505ac4036f69c646dcd43be7a7e8c14ed381bd3dd8e665fec33390485b", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "4b8c549eff80a0d45e732a7d77a1c18318645e3b3d7d62d7c937a2f883ef1003", "generation_seconds_batch": 21.09280116390437, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.425988674163818, "target_z": 1.4218749444139498, "ce": 6.0305562019348145, "genuineness": 4.319066047668457, "identity": 0.392344206571579, "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_DRAM_reg/1f01c1505ac4036f69c646dcd43be7a7e8c14ed381bd3dd8e665fec33390485b.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "1f12ad4f36ef6b6c3bed7a9080c70ccaef3f1c0dc9383d28df95898b6e1a3d13", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "a539679f3d64640125929d2aec077d292ffaa8e65ccaf4ec0b1491551d1fa24b", "generation_seconds_batch": 20.740291732014157, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6471612453460693, "target_z": 0.7382813096732597, "ce": 5.974957466125488, "genuineness": 4.281409740447998, "identity": 0.46464312076568604, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/1f12ad4f36ef6b6c3bed7a9080c70ccaef3f1c0dc9383d28df95898b6e1a3d13.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": "1f8b27b794a2fd4d876e239ac6deb57a2b91725d1b1a961861f7a696dc067fb9", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "7ae9a8f1a1e5298d0acef4949151b3a91b7e765d2ae17cacb7468563454d8725", "generation_seconds_batch": 15.209956906037405, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5180983543395996, "target_z": 0.6250000261581412, "ce": 5.998785495758057, "genuineness": 4.300237655639648, "identity": 0.5383609533309937, "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_DRAM_reg/1f8b27b794a2fd4d876e239ac6deb57a2b91725d1b1a961861f7a696dc067fb9.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "1fa8f54de02d7a0cf55a8659864019084b2acb6c97f1dd445e74086b12f75c49", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "8ca85800eade3622e95a4a982ed7ede5faca907909855bbf3f2a07fc199d6952", "generation_seconds_batch": 16.113950761035085, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.918638229370117, "target_z": 0.9765625147139545, "ce": 5.6334228515625, "genuineness": 4.319066047668457, "identity": 0.275234580039978, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/1fa8f54de02d7a0cf55a8659864019084b2acb6c97f1dd445e74086b12f75c49.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "2004b03f1a597305e78b6d001811179d403e83534c2924a52835fc98c1a29417", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "94ea6a24b2b4cb1e46131b7827643ef83f98c6ac08d4809150cd7273e0ac8e27", "generation_seconds_batch": 22.18088912894018, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3520822525024414, "target_z": -0.3984374591279043, "ce": 6.252951145172119, "genuineness": 4.582659721374512, "identity": 0.6633440852165222, "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_DRAM_reg/2004b03f1a597305e78b6d001811179d403e83534c2924a52835fc98c1a29417.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": "205a642877bfc9a78b8f5701e41ac705921ce7bf5c06d3d7174736167dfe7a01", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "307d44a55f2963caaedc7bc3669dbb6769960a39864a8a53a421d7a556704858", "generation_seconds_batch": 17.580767795909196, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9809443950653076, "target_z": 1.0312499803813941, "ce": 5.242246150970459, "genuineness": 4.080576419830322, "identity": 0.27748924493789673, "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_DRAM_reg/205a642877bfc9a78b8f5701e41ac705921ce7bf5c06d3d7174736167dfe7a01.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "20c6dbe917df23dd81c1b5e6667fc50b03c77a27e8f2025e3ff5a47be219a1f3", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "2c72c8dda24382e13a8d9c8b3e83e6acf68a952096a14a688a1fb6b5a7ff8ee0", "generation_seconds_batch": 21.057567595969886, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8758437633514404, "target_z": 0.06127931725995772, "ce": 5.919358730316162, "genuineness": 4.281409740447998, "identity": 0.7741495370864868, "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_DRAM_reg/20c6dbe917df23dd81c1b5e6667fc50b03c77a27e8f2025e3ff5a47be219a1f3.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "20df08e53de5257baf04c892b9d28dc19d5541a34a2fc73155a85acb36902bfc", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "15d5a8df413d3f26d1ae9cd24c6358c22d02356b1ddfe11a278bacc7c7873172", "generation_seconds_batch": 20.54670853400603, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8405182361602783, "target_z": 0.030273340020051786, "ce": 6.292664527893066, "genuineness": 4.595211505889893, "identity": 0.40505582094192505, "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_DRAM_reg/20df08e53de5257baf04c892b9d28dc19d5541a34a2fc73155a85acb36902bfc.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "20e683e5a1241ef6db60985ecf22266919bdcb45a1c2b92d7b136a3e5e8e09cf", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "f90b3b6dd5529160b2180b99b9d36e83317cba5ddc635816a998722b230f4ce5", "generation_seconds_batch": 21.34764486597851, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.657411575317383, "target_z": 1.6249998168930113, "ce": 5.911416053771973, "genuineness": 3.9173994064331055, "identity": 0.20371869206428528, "hypothesis": "The door is still open. I'd put your coat on the chair. Tell me when you are ready.", "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_DRAM_reg/20e683e5a1241ef6db60985ecf22266919bdcb45a1c2b92d7b136a3e5e8e09cf.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "212263243acdde98f30f4e8fbdf7946ed592a9be73527ba6ab60bb2e553d85d3", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "c89a9543671ceba70e10f0f441d31b38b61a482b0dc8999f7fc7461bec92d110", "generation_seconds_batch": 20.789070877013728, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9873828887939453, "target_z": 0.15917961372586728, "ce": 6.324435234069824, "genuineness": 4.406930446624756, "identity": 0.429533451795578, "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_DRAM_reg/212263243acdde98f30f4e8fbdf7946ed592a9be73527ba6ab60bb2e553d85d3.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": "2168d62c6680684c6e372b707b9895832ff39ff7c4eb38f30c6614e73fcff698", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "3d75aabdbcae30067c94c0e764f8ae1fe188058a2d35820a55ba92c4258ed5df", "generation_seconds_batch": 13.52541666093748, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.731719493865967, "target_z": 0.8124999084465057, "ce": 5.847874641418457, "genuineness": 4.532451152801514, "identity": 0.7824597954750061, "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_DRAM_reg/2168d62c6680684c6e372b707b9895832ff39ff7c4eb38f30c6614e73fcff698.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "21d1d7290b807d4c8bcf292a5ebb5811a059dd1b2ca2c9d0b4cad0356c6c56c7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "0fae28e6861fc880b5bcd75d44ee974812023f8c1cc98e18f200ffa853c3a2e3", "generation_seconds_batch": 20.001645853044465, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4513416290283203, "target_z": 0.5664062083104624, "ce": 6.006728172302246, "genuineness": 4.758388519287109, "identity": 0.46498197317123413, "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_DRAM_reg/21d1d7290b807d4c8bcf292a5ebb5811a059dd1b2ca2c9d0b4cad0356c6c56c7.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "21ec62d40293f71d5624ed1564add048b8242969e0036122cf2aab1eebd99a80", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "6d364c2235893d9e2b4c834de1e6d052b399f6816c7523a3c6a301fe991c7ae5", "generation_seconds_batch": 16.19155024900101, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.286675214767456, "target_z": 0.42187494441394985, "ce": 6.173524379730225, "genuineness": 4.6203155517578125, "identity": 0.6208276748657227, "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_DRAM_reg/21ec62d40293f71d5624ed1564add048b8242969e0036122cf2aab1eebd99a80.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "220493a712dda09397291f9ec258aa6ee245efa7d33e984c350b0d12b84bcb3c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "59fa4b542c15bd88fd40099b7bcabba17981a840520843e0d328d2edbbba75c7", "generation_seconds_batch": 21.252354755066335, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2466213703155518, "target_z": 0.3867188211174465, "ce": 5.744620323181152, "genuineness": 4.036643981933594, "identity": 0.4481421113014221, "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_DRAM_reg/220493a712dda09397291f9ec258aa6ee245efa7d33e984c350b0d12b84bcb3c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "2233094d12e5c0ed3f2c076a4956e907f00558d24d925a7678740e1ef4eaf18d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "d927834e964ec5eaa13e36854721c0c20446a77933d224cdec2e6991d000cf99", "generation_seconds_batch": 16.424036265001632, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.434889793395996, "target_z": 1.4296876487744283, "ce": 5.657250881195068, "genuineness": 4.118232727050781, "identity": 0.2557409107685089, "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_DRAM_reg/2233094d12e5c0ed3f2c076a4956e907f00558d24d925a7678740e1ef4eaf18d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "22554febee4d1c457226f0e0202dd95677cdc0e807a35df8daf282fcfefdfbfd", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "96da91957788ae76bc790ca52fa0635af6625610205c35a9b18228991e90cbe7", "generation_seconds_batch": 21.252354755066335, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.069953441619873, "target_z": 1.1093751406000092, "ce": 5.9431867599487305, "genuineness": 4.293961524963379, "identity": 0.5636681318283081, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/22554febee4d1c457226f0e0202dd95677cdc0e807a35df8daf282fcfefdfbfd.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "2270fde06a6b7a66581531be55cef57d5da3aa49b38ec4c2e94ef0a833f4fd70", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "aa9431484a9744c001ad327608f233da4a67da8ab4974d3567397155b08e8dc3", "generation_seconds_batch": 21.62540798506234, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.732595205307007, "target_z": -0.06445303222034279, "ce": 6.268836498260498, "genuineness": 4.18099308013916, "identity": 0.12298864871263504, "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_DRAM_reg/2270fde06a6b7a66581531be55cef57d5da3aa49b38ec4c2e94ef0a833f4fd70.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "22c3d9d9433a6b20f78b666e8515006174340b50a0485dcb3d4d32bc755fad89", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "f929539238ddab860f7ab5a37d4e0bac0dc955868afde203b602182b426870d0", "generation_seconds_batch": 19.633071449003182, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.586204528808594, "target_z": 1.5624998561302232, "ce": 5.510311126708984, "genuineness": 3.8546390533447266, "identity": 0.30914944410324097, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude graam. We kunnen er 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_DRAM_reg/22c3d9d9433a6b20f78b666e8515006174340b50a0485dcb3d4d32bc755fad89.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "2321d8aa446122582d66c05ea1270bf43008da1f93ca05c6d18b06823d9eb0cd", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "b77999f8cc8b59f3716740a56899c8e42fe84fe6cc45765a3a952fc29cf11b3a", "generation_seconds_batch": 21.874433964956552, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.825178861618042, "target_z": 0.8945312115802301, "ce": 5.5817952156066895, "genuineness": 4.319066047668457, "identity": 0.8184994459152222, "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_DRAM_reg/2321d8aa446122582d66c05ea1270bf43008da1f93ca05c6d18b06823d9eb0cd.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "23c96a2867593bc22c87c54b5db30b8414a2b0200e292e73b80e25cfc82a1944", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "1adfa8caeb83e20d43c558a8d8cb6510c18814aac0a7b3a10e07e43fd2bb8971", "generation_seconds_batch": 21.75819694099482, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.820728302001953, "target_z": 0.8906248593999908, "ce": 5.7247633934021, "genuineness": 4.193545341491699, "identity": 0.8764715194702148, "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_DRAM_reg/23c96a2867593bc22c87c54b5db30b8414a2b0200e292e73b80e25cfc82a1944.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "2531d53121c0ddc466aa6f1893c2f8b1348fc7666c8b383de87652cac1d96d19", "bucket": "vn_S_DRAM_reg", "phase": "wording", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "265c83883bf57ed03142e6bfc54e81c32d9d8037dbb3777e0d2bcd5fb9d4311b", "generation_seconds_batch": 16.447103018057533, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6146583557128906, "target_z": -0.16796874918255808, "ce": 6.149696350097656, "genuineness": 4.168440818786621, "identity": 0.5024688243865967, "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_DRAM_reg/2531d53121c0ddc466aa6f1893c2f8b1348fc7666c8b383de87652cac1d96d19.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "258b6bb204b531b843d6449f3825b98b4fbac4ff859d17385e01e5d86146c5ec", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "3673d2afd1c6176f6f020e5c3134f3c20e6664c1b192261e712c449200cfc77f", "generation_seconds_batch": 20.3628170019947, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1976664066314697, "target_z": 0.3437499934604647, "ce": 6.046441555023193, "genuineness": 4.532451152801514, "identity": 0.5911283493041992, "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_DRAM_reg/258b6bb204b531b843d6449f3825b98b4fbac4ff859d17385e01e5d86146c5ec.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": "25b06369447e19116bba7e71ca3f015a54461f75c0cda996a814a1f6dd95f4bb", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "d6041798bc4d34539f0d6578b212e13fdbb133902b496363f77ca610d012ed2d", "generation_seconds_batch": 15.764855118002743, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.478044271469116, "target_z": 0.5898436935965079, "ce": 6.117925643920898, "genuineness": 4.3504462242126465, "identity": 0.5210590362548828, "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_DRAM_reg/25b06369447e19116bba7e71ca3f015a54461f75c0cda996a814a1f6dd95f4bb.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "25d3e565f29ad268de83d2c79a08ffc70f4fbead124b78acf69a8b6a62e515e7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "26eee680e2d4eeecf409683a9bc7ada2d6297363e5928bfdfa8f4386dbe24df8", "generation_seconds_batch": 21.200859484961256, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.682527542114258, "target_z": -0.10839850023866689, "ce": 6.173524379730225, "genuineness": 4.595211505889893, "identity": 0.6619236469268799, "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_DRAM_reg/25d3e565f29ad268de83d2c79a08ffc70f4fbead124b78acf69a8b6a62e515e7.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "f2d4cbf659abbdaa57e33f8500c679781a85fedf4d9e6724d4109d63a8220fa5", "generation_seconds_batch": 15.31716720201075, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8847856521606445, "target_z": -0.8085937655313964, "ce": 6.348263263702393, "genuineness": 4.243753433227539, "identity": 0.20789240300655365, "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_DRAM_reg/26d12995a5b024b42cbe037fdee48ffb67b79df656c09f059592185aa881aca9.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "27b33f61b0c781bf582324976d143af1ed134e6633633db19ebc13d9f545aaeb", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "9d2949fc82699fe547f366d9df70d2c707bdd8219998f7aecec5c1e5e44acdac", "generation_seconds_batch": 18.531556913047098, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.998746156692505, "target_z": 1.046874970572091, "ce": 5.855817794799805, "genuineness": 4.105680465698242, "identity": 0.4529772400856018, "hypothesis": "Ik heb een klein doosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.1111111111111111, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_DRAM_reg/27b33f61b0c781bf582324976d143af1ed134e6633633db19ebc13d9f545aaeb.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "27be10243243a7af59b32303643dffec812c9db1193bdbf2664f01c3d063c767", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "5dc0e51d7149b645361dba53a99f686484a6a9632fef5a44fd20699f122535ff", "generation_seconds_batch": 21.156741929939017, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.557915210723877, "target_z": -0.21777343157354612, "ce": 6.1894097328186035, "genuineness": 4.482243061065674, "identity": 0.6465146541595459, "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_DRAM_reg/27be10243243a7af59b32303643dffec812c9db1193bdbf2664f01c3d063c767.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "2810e7b914dae5d4d05556b86a34fdff720e7d1fca29e292fb176d09bc25c359", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "b7b780ba3053174e0a330e757bec9959f602fe3b03dd77b602ab3207c7704160", "generation_seconds_batch": 15.901571974041872, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.069953441619873, "target_z": 1.1093751406000092, "ce": 5.5539960861206055, "genuineness": 4.18726921081543, "identity": 0.5384145975112915, "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_DRAM_reg/2810e7b914dae5d4d05556b86a34fdff720e7d1fca29e292fb176d09bc25c359.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "28fb4e9872c422d36854a0b5171364b45cd5ba8f280302e5c536c65652ed7c9e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "1edc8f216877ee0b4e096b64ed768a5165f01f77e5477f60ca28cb330d26b060", "generation_seconds_batch": 22.301188891986385, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.235495090484619, "target_z": 0.3769530452994134, "ce": 6.292664527893066, "genuineness": 4.582659721374512, "identity": 0.3418866991996765, "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_DRAM_reg/28fb4e9872c422d36854a0b5171364b45cd5ba8f280302e5c536c65652ed7c9e.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "28ffc80f6e71ff09f8c0e439bb24108a2234f72b0e1ec123982d1c96d368d6f1", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "a4ad911b33985cd5aa76e12f89d83e2333a00095ea5b93a8292659fe0f0420e3", "generation_seconds_batch": 14.972811780055054, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.777099609375, "target_z": -0.0253905567436002, "ce": 6.133810997009277, "genuineness": 4.545003414154053, "identity": 0.6328176259994507, "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_DRAM_reg/28ffc80f6e71ff09f8c0e439bb24108a2234f72b0e1ec123982d1c96d368d6f1.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "290405a12fbbc576f1503a8b1dcd4de1e4d8edcf2d2a2766258734daa03672e4", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "a3914cb3b2f6d890c9ee833c05c53a2be4cd6041962d8c0991d096703668dc9d", "generation_seconds_batch": 16.126302730990574, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8340797424316406, "target_z": 0.9023437066755786, "ce": 5.462655067443848, "genuineness": 4.243753433227539, "identity": 0.4247824549674988, "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_DRAM_reg/290405a12fbbc576f1503a8b1dcd4de1e4d8edcf2d2a2766258734daa03672e4.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "295e7254b6fd640e5a4b1a1de36169f9fb844c9c5923a52f55d5a00eb97e08ab", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "2cfc155148537ad854935a14120ac301cd27c5188ebeb621399017f6c50d0c4c", "generation_seconds_batch": 20.87778265099041, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.946216344833374, "target_z": 0.12304685006802163, "ce": 6.173524379730225, "genuineness": 4.18099308013916, "identity": 0.2562449872493744, "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_DRAM_reg/295e7254b6fd640e5a4b1a1de36169f9fb844c9c5923a52f55d5a00eb97e08ab.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "2a50e35040e9d79c5dccd3e3e5ca2bdfe4fd90d2e4a285ecedf7d0462056ac93", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "3bcdebe1709e084ada7384fd1178366cda5ca1f4864f1aa528c63eceb96b6164", "generation_seconds_batch": 20.212906358996406, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.411287784576416, "target_z": 0.531250085013959, "ce": 6.0305562019348145, "genuineness": 4.532451152801514, "identity": 0.47616875171661377, "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_DRAM_reg/2a50e35040e9d79c5dccd3e3e5ca2bdfe4fd90d2e4a285ecedf7d0462056ac93.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "2a707000bc8369901d9e8c95929d5e6fe19b9e8dc64e581eeab50b2977efa5a1", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "23b1d186ccfb22a6facb50784c8060ccb8e796ed2eac295856083da995e12421", "generation_seconds_batch": 21.195769233978353, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.678314208984375, "target_z": 0.7656249378744145, "ce": 5.982900142669678, "genuineness": 4.406930446624756, "identity": 0.5482823848724365, "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_DRAM_reg/2a707000bc8369901d9e8c95929d5e6fe19b9e8dc64e581eeab50b2977efa5a1.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "2aec400dafd0612ffc3717bc0c9d03a7f498fad96a67b2ef9b095cb37a3e9b49", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "5d7ec7b86f15d1161d876175c2c8c8a680c3faef8214d39d9fdc3e9d8f54bf99", "generation_seconds_batch": 15.744220194988884, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8234119415283203, "target_z": 0.015258776213334917, "ce": 6.133810997009277, "genuineness": 4.545003414154053, "identity": 0.7550360560417175, "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_DRAM_reg/2aec400dafd0612ffc3717bc0c9d03a7f498fad96a67b2ef9b095cb37a3e9b49.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": "2b32a1ec9aed9480606e7076aced76b3826691dec2ec6a8e52ff36ef00020471", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "861689b16f9ac50fd5b0ba4a3f08b3da4bf9bb4dfecaf1b1851622630385effe", "generation_seconds_batch": 21.890564540051855, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7628726959228516, "target_z": 0.8398437459127904, "ce": 5.732706069946289, "genuineness": 4.093128204345703, "identity": 0.3924160897731781, "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_DRAM_reg/2b32a1ec9aed9480606e7076aced76b3826691dec2ec6a8e52ff36ef00020471.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "2b76fc849590f7094b32dda7b5c45b98931c97ae607e6833e6b479d6bba4ae3c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "915818c7f59edd0ad7138ed21b7de9c7b27eb057de92bb6482dfd1a10ed9f8ff", "generation_seconds_batch": 15.744220194988884, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4201884269714355, "target_z": 0.5390623708441776, "ce": 6.046441555023193, "genuineness": 4.507347106933594, "identity": 0.665012001991272, "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_DRAM_reg/2b76fc849590f7094b32dda7b5c45b98931c97ae607e6833e6b479d6bba4ae3c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "2cb5c7fa08f381875f162dd3a1560d3e3f616c14efa88eadeb5665128381a408", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "71e47f0d67708443dbacd9a1fb8b05d490b8230c6c3872e48765360d77b2fc2a", "generation_seconds_batch": 21.117827340960503, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.731719493865967, "target_z": 0.8124999084465057, "ce": 5.967014789581299, "genuineness": 4.457139015197754, "identity": 0.17310751974582672, "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": 0.03999999999999915, "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_DRAM_reg/2cb5c7fa08f381875f162dd3a1560d3e3f616c14efa88eadeb5665128381a408.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "2cdce6a4494eba8c280928e462ca9fc16b2814569479c5b587b942cac5a73b54", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "ea788464892f7c653bd14e73589e0edffa3a3fca5a9e414274c6d64a6909e092", "generation_seconds_batch": 20.985833944985643, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.408186912536621, "target_z": 1.406249954223253, "ce": 6.09409761428833, "genuineness": 4.337893962860107, "identity": 0.44816312193870544, "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_DRAM_reg/2cdce6a4494eba8c280928e462ca9fc16b2814569479c5b587b942cac5a73b54.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": "2d2fa2db32a355bc0eb97d454b0f67b9f0dc738f111babcdd0d85114a0ab6e62", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5876210a4f8d325df97f60b68b4ae371a8975864f9e6e0af7ab90f7c7be4d8af", "generation_seconds_batch": 20.590041282004677, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6160080432891846, "target_z": 0.7109374722069749, "ce": 6.181467056274414, "genuineness": 4.425758361816406, "identity": 0.5508143305778503, "hypothesis": "Ich schrieb die Nummer auf den Umschlag, ließe genau und prüfe die letzten beiden Ziffern. Ruf an, wenn du Zeit hast.", "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_DRAM_reg/2d2fa2db32a355bc0eb97d454b0f67b9f0dc738f111babcdd0d85114a0ab6e62.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "2da2966e3efba24d8ba78159ac2158d6277ef419737aa73537c4446c5cdd1025", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "e37fb0b8cf017315064e15e0c2ae5d421ea3a80ebf4d7184c77a4b9f358c77dc", "generation_seconds_batch": 20.28390231996309, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.577303886413574, "target_z": 1.5546875703000045, "ce": 5.903473377227783, "genuineness": 3.6663577556610107, "identity": 0.5644485354423523, "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_DRAM_reg/2da2966e3efba24d8ba78159ac2158d6277ef419737aa73537c4446c5cdd1025.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "2e6a2ee935d205ef2343be211d3150d542e4d3cbf2d3cad83c7f562704bcdc20", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "87040dd20e5dacf2272113a73f33bddf4a064dfb5f122569d63a0d44ea1c8d0a", "generation_seconds_batch": 21.436159347067587, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5990817546844482, "target_z": -0.18164066791570047, "ce": 6.054384231567383, "genuineness": 4.7081804275512695, "identity": 0.7736473083496094, "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": 0.03999999999999915, "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_DRAM_reg/2e6a2ee935d205ef2343be211d3150d542e4d3cbf2d3cad83c7f562704bcdc20.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "2ef1f0a6ee3b05b3d05fa012e72310705f4161d2f3e8ea9361d7d2d6448e68a9", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "71e21064ef30be6abffaf4e0199f0932393952fa9cf7c56215b17e95f7985d00", "generation_seconds_batch": 20.576529294019565, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.851881504058838, "target_z": 0.9179686968662756, "ce": 5.315715789794922, "genuineness": 4.406930446624756, "identity": 0.5700333118438721, "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_DRAM_reg/2ef1f0a6ee3b05b3d05fa012e72310705f4161d2f3e8ea9361d7d2d6448e68a9.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "305925894c9362afbd1e29f0fc91e9429d96e6c5818257904a5e3bfe646293a1", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "f3acf91e713b889debe39694010691c9cfb3adf9abb39892b7764f527d41f359", "generation_seconds_batch": 21.423099903040566, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.406837224960327, "target_z": 0.5273437328337198, "ce": 6.086154937744141, "genuineness": 3.773050308227539, "identity": 0.4368458688259125, "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_DRAM_reg/305925894c9362afbd1e29f0fc91e9429d96e6c5818257904a5e3bfe646293a1.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "30854284ed097803385736b12acdaf2da0ab1d6abed57a4809b8024684d291a5", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "4edf3fc012a4eda33a20994d4b5c3adbb1abd57398c9c45d77f01c6eb0bd3fa6", "generation_seconds_batch": 15.23129603289999, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.621808052062988, "target_z": 1.5937498365116172, "ce": 6.125868320465088, "genuineness": 4.042920112609863, "identity": 0.3141833543777466, "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_DRAM_reg/30854284ed097803385736b12acdaf2da0ab1d6abed57a4809b8024684d291a5.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "30c7ff34d07cd138d4e3e400a7b70257747cb35d890d69e5be56bd9f2df8d011", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "3edea6ae976fe4dd685ebfa641ebd493d744be7de5308e2c4879f86312f45fb9", "generation_seconds_batch": 22.730503169004805, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.025448799133301, "target_z": 1.0703124558581367, "ce": 5.4944257736206055, "genuineness": 4.306513786315918, "identity": 0.243180513381958, "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_DRAM_reg/30c7ff34d07cd138d4e3e400a7b70257747cb35d890d69e5be56bd9f2df8d011.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": "30cec2ab2ecf44494f944c53500ab21016d2a1554314e0748151d3eb4252a472", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5d8ee794247823f731b02bd789dd391bdd886612b39f99a021dd552fbb0e5542", "generation_seconds_batch": 20.403864764026366, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6471612453460693, "target_z": 0.7382813096732597, "ce": 5.772419452667236, "genuineness": 4.482243061065674, "identity": 0.444568932056427, "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_DRAM_reg/30cec2ab2ecf44494f944c53500ab21016d2a1554314e0748151d3eb4252a472.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "312f3ddd0b666be423cc8fc5384b79fd05f9ebee0209c1856f3aa31bd273eb50", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "b4958fa1d2bbc82cef3feb2e6539f11924da73f8cb839d0ad9063a4d117fc4f4", "generation_seconds_batch": 21.150231357081793, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 5.2270684242248535, "target_z": 2.124999921525576, "ce": 5.577824115753174, "genuineness": 3.6349775791168213, "identity": 0.3715951442718506, "hypothesis": "Ik heb het kleine toosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "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_DRAM_reg/312f3ddd0b666be423cc8fc5384b79fd05f9ebee0209c1856f3aa31bd273eb50.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "31c3706360bb5c712a1aeb5560801e2710ba94ad564492cd7faff3dd11f8be05", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "ec285b14e6942ea0072dc11faef17b52a81424623138aa9b201077d128d42b0d", "generation_seconds_batch": 20.97356319206301, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0085225105285645, "target_z": 0.17773431573546125, "ce": 6.062326908111572, "genuineness": 4.1998209953308105, "identity": 0.4778883159160614, "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_DRAM_reg/31c3706360bb5c712a1aeb5560801e2710ba94ad564492cd7faff3dd11f8be05.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "3217fe577b4689ea74bd92d05919ce27262d716ba9b982a9695119a9691bfea5", "bucket": "vn_S_DRAM_reg", "phase": "wording", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "7ce9676b40f2839ed5ce8b5db822f54ff3ec9e53ecf99fb4c1ce206474dce5d4", "generation_seconds_batch": 16.447103018057533, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2319202423095703, "target_z": -0.5039062475476742, "ce": 6.268836498260498, "genuineness": 4.871357440948486, "identity": 0.7517073154449463, "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_DRAM_reg/3217fe577b4689ea74bd92d05919ce27262d716ba9b982a9695119a9691bfea5.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": "32263a00542086b213b4b09b409893d2b76273ac19213679576404fb10186bc8", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1cf099480113d4c1cae976c91d5380310c757efe4ab5ac2dfe5cfc3acd84eee7", "generation_seconds_batch": 21.327835389995016, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.998746156692505, "target_z": 1.046874970572091, "ce": 5.474569320678711, "genuineness": 4.193545341491699, "identity": 0.4061379134654999, "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_DRAM_reg/32263a00542086b213b4b09b409893d2b76273ac19213679576404fb10186bc8.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": "3241fb39efe7352b98cdb4fae73a6d4574a68f4ed252c2ff6c143713d1aae309", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "70ab3175464f8d8a2e7905a29c37f399961e88c5df578c9041827077a3e7bb05", "generation_seconds_batch": 15.839278992032632, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7361700534820557, "target_z": 0.8164062606267449, "ce": 5.728734970092773, "genuineness": 4.193545341491699, "identity": 0.4457632303237915, "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_DRAM_reg/3241fb39efe7352b98cdb4fae73a6d4574a68f4ed252c2ff6c143713d1aae309.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "32e3f6f5ede98d9f75f77a867c286560fd90f1e102ffebf6c8641d3759aeca36", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "b1ff5e796d7df497fce12a6d3c3bd4e28be14269f293c70d56d81064d7d40048", "generation_seconds_batch": 21.286601127008907, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6961159706115723, "target_z": 0.7812499280651116, "ce": 5.510311126708984, "genuineness": 4.356721878051758, "identity": 0.36839431524276733, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo. Sobre la silla, dime cuándo estás listo.", "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_DRAM_reg/32e3f6f5ede98d9f75f77a867c286560fd90f1e102ffebf6c8641d3759aeca36.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "3350e242a8a38d0583dc05100e09c3b44c090410471495f876127eb596acc556", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "ce7535a97a23a114fca99fdefd669b531689b346de056377247c799f099cf0f4", "generation_seconds_batch": 20.709671157994308, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8541479110717773, "target_z": 0.04223639970225757, "ce": 6.062326908111572, "genuineness": 4.469690799713135, "identity": 0.7398765683174133, "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_DRAM_reg/3350e242a8a38d0583dc05100e09c3b44c090410471495f876127eb596acc556.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "336ac79350a7a2ef791bac3e7403edbfcbfd427d58c02dde4d7d6bdeb180ac69", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "7982af6fedc0dc27311552c6e08ac968189926a8108a7b7d61631ccf7ae277da", "generation_seconds_batch": 14.719412734964862, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.150061130523682, "target_z": 1.1796873871930158, "ce": 5.903473377227783, "genuineness": 4.432034492492676, "identity": 0.4661741554737091, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/336ac79350a7a2ef791bac3e7403edbfcbfd427d58c02dde4d7d6bdeb180ac69.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "33a16b8b07d9624f5e1d5dc3dd1798c7d86d962f6d81fbeff4b87cd4088f438c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "ccfb7225a062b00db0ed56ffe74c9b377099dac1ba64ea2e01c7f9d922052b42", "generation_seconds_batch": 18.890396273927763, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7097866535186768, "target_z": -0.08447259013938523, "ce": 6.062326908111572, "genuineness": 4.375550270080566, "identity": 0.6841078400611877, "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_DRAM_reg/33a16b8b07d9624f5e1d5dc3dd1798c7d86d962f6d81fbeff4b87cd4088f438c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "33cff22df28deb0f3c8bceb3b0a6006f491f2f91ceec05940f1948c24a484ec6", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "6b3f3402f268e14be7633c5e233a6a7783713221d4c752e93faee87aa02fda2f", "generation_seconds_batch": 16.027165808947757, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.870558738708496, "target_z": 0.05664053712499427, "ce": 6.109982967376709, "genuineness": 4.758388519287109, "identity": 0.8421653509140015, "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_DRAM_reg/33cff22df28deb0f3c8bceb3b0a6006f491f2f91ceec05940f1948c24a484ec6.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "6769c47b4f790b433e8da8a9358bc303213be049880ca071804026505f886aa0", "generation_seconds_batch": 16.028769574943, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3698840141296387, "target_z": -0.3828124689372073, "ce": 6.070269584655762, "genuineness": 4.331617832183838, "identity": 0.8893109560012817, "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_DRAM_reg/33f63538ac2c49676cc53462aa2024b47e58cb29b8690774a6580785ecb0e60e.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": "3416be307a3423d52eeb434d80d6d8812efe4e1e588ed8d5b2a9ff355b75a833", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "3600736c45f1c5dc9ae43c597d870c9375b87d86cdc6ed330f25b7d14d775833", "generation_seconds_batch": 20.588639173889533, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9898452758789062, "target_z": 1.0390624754767426, "ce": 5.371314525604248, "genuineness": 4.149612903594971, "identity": 0.3391147255897522, "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": 0.03999999999999915, "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_DRAM_reg/3416be307a3423d52eeb434d80d6d8812efe4e1e588ed8d5b2a9ff355b75a833.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "346198dd4021c61463c8d9aa42afd031ff199ab9b2822119cb9dd81f93aeeb96", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "6d8bee462e14a54e236c01240fbafd411c0663b15754b662a0a83cdc6c04b06b", "generation_seconds_batch": 22.337924418970942, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.537888288497925, "target_z": -0.2353514932217978, "ce": 6.173524379730225, "genuineness": 4.21237325668335, "identity": 0.7103996276855469, "hypothesis": "He encontrado la cajita. Esta 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_DRAM_reg/346198dd4021c61463c8d9aa42afd031ff199ab9b2822119cb9dd81f93aeeb96.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "346ce2aee4f6e66598023389208eed0fa7583c9f4c68670bf3c84d8d5c227b33", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "32d0d32743637e14527e20e5fb4e8b4f2a0bc247bcdcd813be3f55352ca4e4e2", "generation_seconds_batch": 20.887976292986423, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7228188514709473, "target_z": 0.804687622616287, "ce": 5.129063129425049, "genuineness": 4.362998008728027, "identity": 0.2946924567222595, "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_DRAM_reg/346ce2aee4f6e66598023389208eed0fa7583c9f4c68670bf3c84d8d5c227b33.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "e94a42852d5b73a3d8c72688912e1ed1fca02c9aa5866609b14046daed324646", "generation_seconds_batch": 13.414194418932311, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0619277954101562, "target_z": 0.22460928630755236, "ce": 6.38003396987915, "genuineness": 3.8922953605651855, "identity": 0.4215128421783447, "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_DRAM_reg/347b084eb1c87968a3fc14b855043a2de140e81cf0874b7d65525285bfb31414.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "348f0faf3b8a219526cdc2e2db467d12e8c6fcacc18dada2c7c324c7ed2383f3", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "6a16a9a5b6d7218175fbcb7d324a7e7681e1e688b03d1965f962b84b0f146168", "generation_seconds_batch": 21.926073814043775, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0875179767608643, "target_z": 0.24707034049738555, "ce": 6.268836498260498, "genuineness": 4.13078498840332, "identity": 0.1988159865140915, "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_DRAM_reg/348f0faf3b8a219526cdc2e2db467d12e8c6fcacc18dada2c7c324c7ed2383f3.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": "349d39999fe089aed3283bd77a13d71858770a19b60c7a94bf6b78ef9b4f1566", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Escribí el número en el sobre.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "f82663b50b222d187b62c5b097839707966ef9f96315ec825329490066c1c849", "generation_seconds_batch": 16.103637556079775, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.123358726501465, "target_z": 1.1562501111721002, "ce": 5.7684478759765625, "genuineness": 4.21237325668335, "identity": 0.5065300464630127, "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_DRAM_reg/349d39999fe089aed3283bd77a13d71858770a19b60c7a94bf6b78ef9b4f1566.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "353399aee17f528393bd5975cec0865069c95582151943f7823fbd1a92b0ac83", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "00d0e4e02e7e30dfb9980aa2e67136efb16843347d8064abb60f1c6cd6806a2b", "generation_seconds_batch": 20.835940978024155, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.918638229370117, "target_z": 0.9765625147139545, "ce": 5.847874641418457, "genuineness": 4.105680465698242, "identity": 0.4476408362388611, "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_DRAM_reg/353399aee17f528393bd5975cec0865069c95582151943f7823fbd1a92b0ac83.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": "354054ee6b21f180922b6c7b1c3cfb422858cdd0edc68cc05034377bf1c1a14d", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Someone moved the little blue chair.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e1ad98761fb6df179d347bba668e6e982098fba3812cc17443cc8b7c7a648d0a", "generation_seconds_batch": 21.23578968003858, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.758422374725342, "target_z": 0.8359376029976812, "ce": 6.078212261199951, "genuineness": 3.9173994064331055, "identity": 0.16769680380821228, "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_DRAM_reg/354054ee6b21f180922b6c7b1c3cfb422858cdd0edc68cc05034377bf1c1a14d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "355b4f33f95d0281435ce5103a95d3749211ea6a81a256ea6ba5cff26fd8729e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "5c131a637bdb9a83d2d473bd4a5bb4e32aac609207c83db80c85d7636bfededc", "generation_seconds_batch": 16.183992760023102, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.924520492553711, "target_z": 0.10400393251032147, "ce": 6.213237762451172, "genuineness": 4.632867813110352, "identity": 0.7729229927062988, "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_DRAM_reg/355b4f33f95d0281435ce5103a95d3749211ea6a81a256ea6ba5cff26fd8729e.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "358798725643137510d102c840640846aab9cb8fab6bf7d20b196e13baa19994", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "fa865695828929660d66738377802ddf8f5217cdfc3931da6b4fd752dec0206b", "generation_seconds_batch": 14.750742144999094, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6961159706115723, "target_z": 0.7812499280651116, "ce": 6.078212261199951, "genuineness": 4.331617832183838, "identity": 0.5154120922088623, "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_DRAM_reg/358798725643137510d102c840640846aab9cb8fab6bf7d20b196e13baa19994.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "365892140104a018ac34a837e02986758739edecef0b0893c0f0eb08022cd202", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "3fde33b055aa18a00e094cac6d527106309cf41fe5ecf0aa03cf2769f7ba36fd", "generation_seconds_batch": 22.18088912894018, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7807154655456543, "target_z": -0.022216841783215136, "ce": 6.109982967376709, "genuineness": 4.582659721374512, "identity": 0.562158465385437, "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_DRAM_reg/365892140104a018ac34a837e02986758739edecef0b0893c0f0eb08022cd202.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "37023e8198c99e5081bf754c90b1998055d57435c796b5a793506d60df1fd5d9", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "682cec2e2b3ae20a54276f3bfea0d936a49a4c5a70fcc8007f22ae0b88db4052", "generation_seconds_batch": 21.567947308998555, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.141160488128662, "target_z": 1.1718751013627973, "ce": 5.677107334136963, "genuineness": 4.13078498840332, "identity": 0.3667088449001312, "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_DRAM_reg/37023e8198c99e5081bf754c90b1998055d57435c796b5a793506d60df1fd5d9.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "3787a0fba3fb91c88e17ed17376aa01c7975ddadcda422749878dbb1b39fd9fe", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "a8612b76ee7489a9f579233bbc5af8a628ada83634159ac75b3a82c7f65ee999", "generation_seconds_batch": 20.8651416299399, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6675074100494385, "target_z": -0.12158199415857317, "ce": 6.300607204437256, "genuineness": 4.519899368286133, "identity": 0.5641404986381531, "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_DRAM_reg/3787a0fba3fb91c88e17ed17376aa01c7975ddadcda422749878dbb1b39fd9fe.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "37bfbc44b741e752762bcbfd246366d1489dd70c8709c61e4889e7153211c542", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "dd81a7364efcfb2400d7ed036fa8a02dece7d53a4175412f9f24b45174655b54", "generation_seconds_batch": 19.693563613924198, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8385300636291504, "target_z": 0.9062498495906879, "ce": 6.308549880981445, "genuineness": 4.432034492492676, "identity": 0.6169170141220093, "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_DRAM_reg/37bfbc44b741e752762bcbfd246366d1489dd70c8709c61e4889e7153211c542.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "3815913691e8a79bcfe22b61b925bb6296e54c3a29ade63e20fed46490673aa8", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "fc706e097de1f4dbe9712f4ebf65ec2343d35dee3d87e082312263fa6e88070d", "generation_seconds_batch": 21.68604837998282, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8118276596069336, "target_z": 0.8828125735697723, "ce": 5.2918877601623535, "genuineness": 4.432034492492676, "identity": 0.40912991762161255, "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_DRAM_reg/3815913691e8a79bcfe22b61b925bb6296e54c3a29ade63e20fed46490673aa8.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "387a5cb8452df296f659283307182f09c4a055d765ff38b22f469a5bda5ffee3", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "34a4803313bf326db17a749e07358eb554c089eba3e6df8920cb17a7266ff65d", "generation_seconds_batch": 19.513697082991712, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7228188514709473, "target_z": 0.804687622616287, "ce": 5.7684478759765625, "genuineness": 4.557555198669434, "identity": 0.7453380227088928, "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_DRAM_reg/387a5cb8452df296f659283307182f09c4a055d765ff38b22f469a5bda5ffee3.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "396f05ec732ece768f41a1a940757a4ccfb4e670394352b355fb2200e4eba3e2", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "a068d2eb0867d9e89155b811e64ed6913856105c9f5b93d585f644a7f001fbab", "generation_seconds_batch": 21.181547805899754, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.185664653778076, "target_z": 1.21093736757441, "ce": 5.589737892150879, "genuineness": 4.337893962860107, "identity": 0.6422896981239319, "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_DRAM_reg/396f05ec732ece768f41a1a940757a4ccfb4e670394352b355fb2200e4eba3e2.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": "3b307406c6bd53dd4be66234a33254c67a6d8c769af8a86b7214c725b3846d42", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "fb3fae4e720ec69a04ec7aca047beafb1ecd4e8d165608b1f8a9a46407bc70f1", "generation_seconds_batch": 20.590041282004677, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.105556964874268, "target_z": 1.1406251209814033, "ce": 5.935244083404541, "genuineness": 4.319066047668457, "identity": 0.5277882814407349, "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_DRAM_reg/3b307406c6bd53dd4be66234a33254c67a6d8c769af8a86b7214c725b3846d42.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "3b6d5be976dc2f5c4cf91ebb985bbdcba8bbb7e9867951d21d28c262364a50f5", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "0042ca87893b808d331c656db46324add78ea930910cfcccec63b13ab49a8081", "generation_seconds_batch": 16.488415655097924, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0574774742126465, "target_z": 0.22070314339244307, "ce": 6.125868320465088, "genuineness": 3.9173994064331055, "identity": 0.5279419422149658, "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_DRAM_reg/3b6d5be976dc2f5c4cf91ebb985bbdcba8bbb7e9867951d21d28c262364a50f5.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "3b72d07cbcc74ae73959ba8fb374e9c056bba054b743d5eff1c0ccff438ef9b2", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "036ba915f21cf8cffe01031d8c63c41f2ef6811c7967ff6e82ac0b8080f3f0e4", "generation_seconds_batch": 20.80735570890829, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.256872177124023, "target_z": 1.273437746867458, "ce": 5.895530700683594, "genuineness": 4.15588903427124, "identity": 0.16664919257164001, "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_DRAM_reg/3b72d07cbcc74ae73959ba8fb374e9c056bba054b743d5eff1c0ccff438ef9b2.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": "3bd04d605ba7944c1bbb00b1f8edcc8c9f40f074a2b61570340adfa9430fb972", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The road turns left behind the bridge.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "26aaa41841943f18711b34560a98109214666ab6e5cdab09f932cc85e8d87933", "generation_seconds_batch": 19.970064062974416, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.559502124786377, "target_z": 1.5390625801093076, "ce": 6.070269584655762, "genuineness": 3.810706615447998, "identity": 0.4266532361507416, "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_DRAM_reg/3bd04d605ba7944c1bbb00b1f8edcc8c9f40f074a2b61570340adfa9430fb972.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "3c0c6f5ce89cddc5da4e15469aede2ecf883ec8b3ed34e605cc07dc4ec86c20d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "f49461bb18361ae079f1244eb4c90512f4800afac8607af2a248554df8db67be", "generation_seconds_batch": 15.622418796061538, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5448009967803955, "target_z": 0.6484375114441868, "ce": 5.669164657592773, "genuineness": 4.482243061065674, "identity": 0.24640943109989166, "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_DRAM_reg/3c0c6f5ce89cddc5da4e15469aede2ecf883ec8b3ed34e605cc07dc4ec86c20d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "3c36ce4f09978be99595cc552a44b56fc5125ff2f2e4e898693f2a4580a7a6e3", "bucket": "vn_S_DRAM_reg", "phase": "wording", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "781d91ea71b5fae04917d227cdee011a1d9bd1e9cb7d9496509ab70b9a3f41ca", "generation_seconds_batch": 14.779266525059938, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2741994857788086, "target_z": -0.4667968435284863, "ce": 6.348263263702393, "genuineness": 4.124508857727051, "identity": 0.567773699760437, "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_DRAM_reg/3c36ce4f09978be99595cc552a44b56fc5125ff2f2e4e898693f2a4580a7a6e3.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "3c3b7c1b38ed93e3689236c31d42c9d8c1a8a49c0db3b789c5d2656927886e5f", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "52f441aaad872bea2866abf4dc6b9181945c172c97074a34d16de3cdec2bc85e", "generation_seconds_batch": 19.48480958910659, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4513416290283203, "target_z": 0.5664062083104624, "ce": 4.90170431137085, "genuineness": 4.306513786315918, "identity": 0.4433538019657135, "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_DRAM_reg/3c3b7c1b38ed93e3689236c31d42c9d8c1a8a49c0db3b789c5d2656927886e5f.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "ebb642ca2afd62144af85815be7daf023ee9202c21352246f94f750b644a3f4e", "generation_seconds_batch": 15.823777557932772, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3721091747283936, "target_z": -0.38085939747965264, "ce": 6.2767791748046875, "genuineness": 4.174716949462891, "identity": 0.6712387800216675, "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_DRAM_reg/3c590f9302048ad206a88d3c2440f1028cc9eea1f1662250baf365be21a6ebfa.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "3cd6407d5be6dd4581d2cd9a276e9616c70fa43c468d5eacedb132df6fa57913", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "4c9c2ced127391c4a354ddb70834dcc49ff998174ad49682c7f14a16a045cd97", "generation_seconds_batch": 21.553491873899475, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.158962249755859, "target_z": 1.1875000915534943, "ce": 5.577824115753174, "genuineness": 4.143336772918701, "identity": 0.4755099415779114, "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_DRAM_reg/3cd6407d5be6dd4581d2cd9a276e9616c70fa43c468d5eacedb132df6fa57913.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "3d048c9ddc453c9fb4e5495d1bcf28ae1b28afd96bc7c0492376ee43b8f589cf", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "3c016f2db278b81a20dcbd17f94467cfc87c883fdac6135df8fc0338660b7517", "generation_seconds_batch": 21.76063445804175, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5982062816619873, "target_z": 0.6953124820162779, "ce": 6.086154937744141, "genuineness": 4.118232727050781, "identity": 0.32900699973106384, "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_DRAM_reg/3d048c9ddc453c9fb4e5495d1bcf28ae1b28afd96bc7c0492376ee43b8f589cf.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "3d6fdd4ac4c608a181c78db4146c393b31c1c541bde63e297c1148fa5e0ee68f", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "c9489f2110bb0a2c6331f8872e5d9a3d2cb4969e7fbf913f4bd29c612d58a828", "generation_seconds_batch": 20.911640733946115, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.829629421234131, "target_z": 0.8984375637604692, "ce": 5.510311126708984, "genuineness": 4.369274139404297, "identity": 0.5781327486038208, "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_DRAM_reg/3d6fdd4ac4c608a181c78db4146c393b31c1c541bde63e297c1148fa5e0ee68f.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "3f34a5c9036c851feda316ca8c01bc417b85360a405ce37de8ebba04d983db32", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "3903d6ad4d3ec38b7c9b87f0e9ee97ee0ac77fb89b55d2bbe0469062342b1c58", "generation_seconds_batch": 15.804306306992657, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.606870174407959, "target_z": -0.17480460391656433, "ce": 6.165581703186035, "genuineness": 4.6705241203308105, "identity": 0.4142308831214905, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.42105263157894735, "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_DRAM_reg/3f34a5c9036c851feda316ca8c01bc417b85360a405ce37de8ebba04d983db32.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "3f36771e779157d260bb879868f516b4a52a26813f7d9181001f000b9607f671", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "ce3f173616e10f5c48bc3b02d0a37911f108e6d93f47729f617414b12b9640ce", "generation_seconds_batch": 22.577014012029395, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7125680446624756, "target_z": -0.08203130313372439, "ce": 6.403861999511719, "genuineness": 4.595211505889893, "identity": 0.5403473973274231, "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_DRAM_reg/3f36771e779157d260bb879868f516b4a52a26813f7d9181001f000b9607f671.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "3fb2f49ae50d837c64953e357609021338059705154dbbd80a91f806aef100c2", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "e30f1fd9e5d7b112687d9b6897960ca2b391a2432c24ef8c5c86c21ae7f21292", "generation_seconds_batch": 21.37828482792247, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.998746156692505, "target_z": 1.046874970572091, "ce": 5.959072113037109, "genuineness": 4.193545341491699, "identity": 0.3773347735404968, "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_DRAM_reg/3fb2f49ae50d837c64953e357609021338059705154dbbd80a91f806aef100c2.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "3fbd3383f926420b5e53750091a8f018332eb5e503e706930f95d62d798963b7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "c38465c3979538b44900c4d459482373722f01945b8f61aa2684dbadd92debb7", "generation_seconds_batch": 21.61246821191162, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.409937858581543, "target_z": -0.3476563456407039, "ce": 6.284721851348877, "genuineness": 4.7834930419921875, "identity": 0.6054481267929077, "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_DRAM_reg/3fbd3383f926420b5e53750091a8f018332eb5e503e706930f95d62d798963b7.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "40324b4b72266752f0088167f5d75ee2e828ee6f090ffcb1777d5b0c1c44f0fc", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "104d8486b507ce97b6e72dea67905d358bf0481376884c0c30b45baf7447183b", "generation_seconds_batch": 21.772093457984738, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.170963764190674, "target_z": 0.32031250817441914, "ce": 6.1894097328186035, "genuineness": 4.306513786315918, "identity": 0.6208028793334961, "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_DRAM_reg/40324b4b72266752f0088167f5d75ee2e828ee6f090ffcb1777d5b0c1c44f0fc.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "420755685cbdbb773d99ec469b2961faa97f86022cb28fa10158908bf50af927", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "d4a197af0648ebe6cc13d5da7826e4f9b3bc54773250bed09610d1a40b0255e0", "generation_seconds_batch": 20.839448106009513, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0864052772521973, "target_z": 0.24609370013604326, "ce": 6.141753673553467, "genuineness": 4.758388519287109, "identity": 0.6050592660903931, "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_DRAM_reg/420755685cbdbb773d99ec469b2961faa97f86022cb28fa10158908bf50af927.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "421086e577b7678b466e63889fe997006e4c8322fe133356006a70c31b551f69", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "aad4485da6dd620dfa3abe6f5839d036feff7eaaffed86322a3c3b1d9cc9f2d5", "generation_seconds_batch": 15.751054301043041, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1553871631622314, "target_z": 0.30664058944127676, "ce": 6.022613525390625, "genuineness": 4.6956281661987305, "identity": 0.676334023475647, "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_DRAM_reg/421086e577b7678b466e63889fe997006e4c8322fe133356006a70c31b551f69.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "426bbb0ead4446836d36efbba8db33ac893c1e68f7c674472aaa191a2308e30e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "65db1d2fbb617652a91a31ec0fa7c167e96dc28a826d146bc6317ce2f3108863", "generation_seconds_batch": 20.86892750300467, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7501187324523926, "target_z": -0.049072254436263806, "ce": 5.935244083404541, "genuineness": 4.65797233581543, "identity": 0.7925319671630859, "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_DRAM_reg/426bbb0ead4446836d36efbba8db33ac893c1e68f7c674472aaa191a2308e30e.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "42dd700f66cf27521acacebb3d214d4c1c3fec3f084f9b5ec9751b52feadbd6d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "f1b248a767b965a14cac009e322c58f611ff4eb550b331f6523640f21203f1a1", "generation_seconds_batch": 20.504979660967365, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.405487537384033, "target_z": -0.35156248855581324, "ce": 6.252951145172119, "genuineness": 4.595211505889893, "identity": 0.8106358051300049, "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_DRAM_reg/42dd700f66cf27521acacebb3d214d4c1c3fec3f084f9b5ec9751b52feadbd6d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "43101ed89c22869a6e520913d69de5401e635ea34ddeef2dfe766990e0ca2035", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "5e29d3ad4a3f9a1a14a51f6e5686f7ec44ed89924c3435990b635839a07f0af7", "generation_seconds_batch": 21.94769762188662, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.265772819519043, "target_z": 1.2812500326976766, "ce": 5.2640886306762695, "genuineness": 3.8295347690582275, "identity": 0.6739709377288818, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Si posé sur le... Si poussé quand il t'en colle, c'est lié, on neutre ça.", "wer": 0.7777777777777778, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_DRAM_reg/43101ed89c22869a6e520913d69de5401e635ea34ddeef2dfe766990e0ca2035.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "43207fec70faea2892f7592dfdaad31c00b0e6a8549ee2606595f94ec47fdad9", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "425729d6aec94b910eb6c07190aaf928c7ea9dfc9db695ba542f2cc381d9b0ee", "generation_seconds_batch": 20.857217872980982, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6135456562042236, "target_z": -0.16894538954390037, "ce": 6.427689552307129, "genuineness": 4.545003414154053, "identity": 0.4699186682701111, "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_DRAM_reg/43207fec70faea2892f7592dfdaad31c00b0e6a8549ee2606595f94ec47fdad9.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "cf1aa8fb63db3e32f73ae9e29400f33aa3ccd850ac71226be6ea7cc319cac2fb", "generation_seconds_batch": 16.028769574943, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3075778484344482, "target_z": -0.4374999346046469, "ce": 6.09409761428833, "genuineness": 4.13078498840332, "identity": 0.783270001411438, "hypothesis": "De weg gaat linksaf naar de brug. Loopt verder tot aan het kleine houten hek. Ik wacht aan de andere kant.", "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_DRAM_reg/434f953004ee3838b31e4f505ba0cb9202b34ab37936979fff50edb64a6a996f.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "df451f1f6020d50ea82cfdcce92a42901da7b0bf4f8c8ec2428b5ce218368720", "generation_seconds_batch": 16.136952575063333, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9373154640197754, "target_z": 0.11523435497267312, "ce": 6.268836498260498, "genuineness": 4.545003414154053, "identity": 0.5863103866577148, "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_DRAM_reg/43d771a5c055e2aea1186dc99f4cdef2b4c6bc8c6b2779ad35a29ca11c75a91c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "4435b8fcb8f7d175862d7da863fe2f6fa86e3c08b2b23c6d99ce117bc39095a3", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "286c594e184de81b26bf494a904050b74ab4284a908e60c3844d77301df1b7ea", "generation_seconds_batch": 20.096152001991868, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.411287784576416, "target_z": 0.531250085013959, "ce": 6.284721851348877, "genuineness": 4.532451152801514, "identity": 0.5496838092803955, "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_DRAM_reg/4435b8fcb8f7d175862d7da863fe2f6fa86e3c08b2b23c6d99ce117bc39095a3.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": "443e43661d9a6df5b36c2b2065c48caf7cb9b7b0ce86e54edd470f2bd0fb9659", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Your message arrived after I had left.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "a24759f6fe30f60a36f9da262cdb7965ea95bbc988eb45aa9cd51696f3cd2fe4", "generation_seconds_batch": 15.548620968009345, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.728618621826172, "target_z": 1.6874997776557994, "ce": 5.911416053771973, "genuineness": 3.8546390533447266, "identity": 0.589634358882904, "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_DRAM_reg/443e43661d9a6df5b36c2b2065c48caf7cb9b7b0ce86e54edd470f2bd0fb9659.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "4447c71be18f2b80ad244fba58c0a18a4fbdbd4f641ceb303dc1d077141fbfb5", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "a92853b3aa5f850331b89ca94892158ead6588661622932773d98f03106ac1be", "generation_seconds_batch": 21.150231357081793, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.470493316650391, "target_z": 1.4609376291558225, "ce": 5.585766792297363, "genuineness": 3.961331605911255, "identity": 0.2999834418296814, "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_DRAM_reg/4447c71be18f2b80ad244fba58c0a18a4fbdbd4f641ceb303dc1d077141fbfb5.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "445ecb4e3373dfdf72422c759e85f8a665091a04da26b8ec47805eb9ea95ee8a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "aa9f0e1b473a54f9483ead61e43b7ffc0786c5aa10c51e41830a478a8705a470", "generation_seconds_batch": 22.83756926190108, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5979692935943604, "target_z": -0.18261709901191284, "ce": 5.982900142669678, "genuineness": 4.432034492492676, "identity": 0.6370092034339905, "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_DRAM_reg/445ecb4e3373dfdf72422c759e85f8a665091a04da26b8ec47805eb9ea95ee8a.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "44e5644199fa68f7e7fb251e9fa024535818cf0cd30d7a89602e299e31e671d1", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "c0320d6e5582a9173e22f6fa787347bad9e2351c0e9103cb832fe76a896e7d87", "generation_seconds_batch": 16.183992760023102, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1843152046203613, "target_z": 0.3320313554500069, "ce": 6.205295085906982, "genuineness": 4.8588056564331055, "identity": 0.7940709590911865, "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_DRAM_reg/44e5644199fa68f7e7fb251e9fa024535818cf0cd30d7a89602e299e31e671d1.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "44fb1fe8f18cf6a47be265ffc6b8c8c45bc0dcddcea5ba80d74cf52c7dec8a9c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "b79e00388a09c85cc86114215708314cf94eef12127669425895b811bf3fcb54", "generation_seconds_batch": 19.693563613924198, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.3013763427734375, "target_z": 1.3125000130790707, "ce": 6.332377910614014, "genuineness": 4.3881025314331055, "identity": 0.4804227948188782, "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_DRAM_reg/44fb1fe8f18cf6a47be265ffc6b8c8c45bc0dcddcea5ba80d74cf52c7dec8a9c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "45906199604ceb480872e320b7eb2aa628bc732eb0a9a0f1826fd3e32eb09771", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "9cf948b41d3c7a8fdf917e910b7a006111d62d4b446da16801536c35acf8e912", "generation_seconds_batch": 21.269453841028735, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7807154655456543, "target_z": -0.022216841783215136, "ce": 6.38003396987915, "genuineness": 4.432034492492676, "identity": 0.4359917640686035, "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_DRAM_reg/45906199604ceb480872e320b7eb2aa628bc732eb0a9a0f1826fd3e32eb09771.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "7e3d5fe038611a23dce474d6b8f169903da6fe27169052b51486bc20e8be13cf", "generation_seconds_batch": 14.901561959064566, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5690412521362305, "target_z": -0.20800786502064295, "ce": 6.324435234069824, "genuineness": 4.149612903594971, "identity": 0.40158936381340027, "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_DRAM_reg/45f9ec8497c8691f9e18d794cf98f47c522bfab041b36df5accabf5c5c456412.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "46776cfc382a9c5b5498aef0a55de3a82402bfe9ea43f3fb438778de2f8f8ecc", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "7286482a625e842dd379e9037e3737387663bb6d31590e9e27ee4fe1dbefd092", "generation_seconds_batch": 15.640359251061454, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3044769763946533, "target_z": 0.4374999346046469, "ce": 6.133810997009277, "genuineness": 4.369274139404297, "identity": 0.42800259590148926, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/46776cfc382a9c5b5498aef0a55de3a82402bfe9ea43f3fb438778de2f8f8ecc.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "4677c5ed779f445999c69b5abe93ad85fbc9da94a26b21443bcb296d5d71989d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "37f8b0ad6d9e9047957162f80c53a33f4b8cc73e9fca7740c89c13fbed431d28", "generation_seconds_batch": 15.197967244079337, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.864161252975464, "target_z": 0.05102532589381845, "ce": 6.229123115539551, "genuineness": 4.6705241203308105, "identity": 0.6248655319213867, "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_DRAM_reg/4677c5ed779f445999c69b5abe93ad85fbc9da94a26b21443bcb296d5d71989d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "467ed57b9c7dc21db701ac143448aba64409aa8031ba6ca02a2b520b19cb3968", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "6c2aaee81f2fad9dd5a63438e40b47ecb0778a8ee2d351cab1d333c25e886ab7", "generation_seconds_batch": 21.37828482792247, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.247971057891846, "target_z": 1.2656250425069795, "ce": 5.270045757293701, "genuineness": 3.911123275756836, "identity": 0.3316011428833008, "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_DRAM_reg/467ed57b9c7dc21db701ac143448aba64409aa8031ba6ca02a2b520b19cb3968.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "46a5371ab5688daa52c8b19347f797c6132e5e6adb74ae76ff969a6ed6d3f90a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "f4ff3f14e5db8278705aac9a786dde2abe16632759fe6f8d7bdc6a6c4857883c", "generation_seconds_batch": 14.483692134032026, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7673230171203613, "target_z": 0.8437498888278997, "ce": 4.969216823577881, "genuineness": 4.331617832183838, "identity": 0.3948429822921753, "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_DRAM_reg/46a5371ab5688daa52c8b19347f797c6132e5e6adb74ae76ff969a6ed6d3f90a.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "46f552104c1c12b3e88c0b0ac789df5d269c636ddff50da6c9813190784fd275", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "1b4aa2df99fb59febe1386aced2b64eeb9ff0b02958a31cca3b150c35ac11619", "generation_seconds_batch": 20.3966804939555, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2110178470611572, "target_z": 0.35546884073605245, "ce": 6.133810997009277, "genuineness": 4.331617832183838, "identity": 0.38279423117637634, "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_DRAM_reg/46f552104c1c12b3e88c0b0ac789df5d269c636ddff50da6c9813190784fd275.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "47085470023acb7d34a6ec1f5b4a99adfa28eb997de3290cae2d03ddf56a4984", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "2dab12a05d874b91edf8becbf1c17d1104995708e2c88d7b70c862c7fceb8ce4", "generation_seconds_batch": 21.033383051049896, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.998746156692505, "target_z": 1.046874970572091, "ce": 6.038498878479004, "genuineness": 3.9801597595214844, "identity": 0.4967927932739258, "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_DRAM_reg/47085470023acb7d34a6ec1f5b4a99adfa28eb997de3290cae2d03ddf56a4984.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "47280881e162a6ed2df48f71c65f7e743ce3f10e4ae3061eadce227d685713ff", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "0e8dd918a5dee9d6829278500a313b0b0e024c047d2dee8b83d06b0a0504cdd9", "generation_seconds_batch": 20.97356319206301, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6605124473571777, "target_z": 0.7499999476837175, "ce": 6.022613525390625, "genuineness": 4.18099308013916, "identity": 0.5848290920257568, "hypothesis": "La puerta sigue abierta. he dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/47280881e162a6ed2df48f71c65f7e743ce3f10e4ae3061eadce227d685713ff.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "357755624ef6568e6b2b220f4a8fabc97947a5b58d6a511852f562405a35d490", "generation_seconds_batch": 16.15020708704833, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4944963455200195, "target_z": -0.27343753760232803, "ce": 6.022613525390625, "genuineness": 4.545003414154053, "identity": 0.7717330455780029, "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_DRAM_reg/47d00b404f769e82194cfb39f2be7a6bd3cc5bfaac4f38356b0fada2abdb9f4d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "47ec48e0509f786ab580fa310482df0b0979bf367bd302c182f0b971f91f06e8", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "913a26421090a9989b0fd0f4b3eaf3180d8f92572bcbbe8abfd92cfde0cfa753", "generation_seconds_batch": 20.46538573107682, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6981041431427, "target_z": -0.0947265815055245, "ce": 6.165581703186035, "genuineness": 4.582659721374512, "identity": 0.593002438545227, "hypothesis": "Die Tür ist noch offen. Ich habe dein Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist.", "wer": 0.05263157894736842, "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_DRAM_reg/47ec48e0509f786ab580fa310482df0b0979bf367bd302c182f0b971f91f06e8.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "48e58a5d4762e5008b49b91ebe908d9b2edcf2a8cf201d32dbd9f9998bcba2c7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "2d8c4ceca00cd1ae0d72719776280c9c0df88ce80ff51021f9027e426fdb09bf", "generation_seconds_batch": 20.51925923593808, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1487114429473877, "target_z": 0.3007811658034829, "ce": 6.0305562019348145, "genuineness": 4.080576419830322, "identity": 0.5548987984657288, "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_DRAM_reg/48e58a5d4762e5008b49b91ebe908d9b2edcf2a8cf201d32dbd9f9998bcba2c7.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "490114042ddf5beb0dffef681ceec2bc5507a315a11241afbee101e58f91ce1c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "99781a5081649e34d90219a51736c8707f78ffaa205a9efad6ab161218505d78", "generation_seconds_batch": 21.647308001993224, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6853091716766357, "target_z": -0.10595700396787613, "ce": 5.990842819213867, "genuineness": 4.319066047668457, "identity": 0.8246572613716125, "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_DRAM_reg/490114042ddf5beb0dffef681ceec2bc5507a315a11241afbee101e58f91ce1c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "4962c41586a629944c3b7420426ca3fb66e01b5e2a1ab784faed26e50f1da857", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "79558430cb6a12b17a80dd2db57dc93d782b961cd8b07a967e3a012fd86d037e", "generation_seconds_batch": 22.145443430985324, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9097373485565186, "target_z": 0.968750019618606, "ce": 5.895530700683594, "genuineness": 4.18726921081543, "identity": 0.3752822279930115, "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_DRAM_reg/4962c41586a629944c3b7420426ca3fb66e01b5e2a1ab784faed26e50f1da857.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "496bbc1f5d8234d1f65e892a8ea1840bd07c629be9d43bfb1d92de353731fbe3", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "6735de9a7229c52d424f381e021bcf828c4b770dbbb7e2477ceb680195c480bc", "generation_seconds_batch": 21.436159347067587, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.144261121749878, "target_z": 0.2968750228883736, "ce": 5.998785495758057, "genuineness": 4.582659721374512, "identity": 0.7196072340011597, "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_DRAM_reg/496bbc1f5d8234d1f65e892a8ea1840bd07c629be9d43bfb1d92de353731fbe3.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "49a47c4f1fe3f72fe05a3be57fa30ed268f04caffebdbf443d9ff934c460e3b1", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "789ebea3961df98a7def4f6ef10c1864727df478e5c3eba00fb1bef2fbe6beb5", "generation_seconds_batch": 16.06255209504161, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.308927536010742, "target_z": 0.44140628678488614, "ce": 6.197352409362793, "genuineness": 4.519899368286133, "identity": 0.3807368278503418, "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_DRAM_reg/49a47c4f1fe3f72fe05a3be57fa30ed268f04caffebdbf443d9ff934c460e3b1.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "e6cceb7a4d7aba2029807e4918428a09c09b92b07af98764c7ac7fd06118fc62", "generation_seconds_batch": 16.491976363002323, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7854440212249756, "target_z": -0.01806648646148779, "ce": 6.117925643920898, "genuineness": 4.293961524963379, "identity": 0.6675674915313721, "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_DRAM_reg/49d468ae6540cc7c1e5b0e6883c25e2b29599458c9de0a5398000eaca62677ab.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "49f4565786b27b123828fd9cf074e1197267a702e7c7aed4f7de882496236900", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "ba30ba837a03984192bfbd149ea44e19d62580a654ab597cff9b54d69945d6ad", "generation_seconds_batch": 20.73278114804998, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.047463893890381, "target_z": 0.21191400793575227, "ce": 6.165581703186035, "genuineness": 4.557555198669434, "identity": 0.8093502521514893, "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_DRAM_reg/49f4565786b27b123828fd9cf074e1197267a702e7c7aed4f7de882496236900.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "4a49645e895173d28bcd2568689200779c7b3d986a47e052f3560f409ad86fb1", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "7a257fbedef2351998eeb7324f70757414c780068fca5fcb0824ce30cc4e44bb", "generation_seconds_batch": 21.82172237301711, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6694133281707764, "target_z": 0.757812442779066, "ce": 5.279973983764648, "genuineness": 4.13078498840332, "identity": 0.555355429649353, "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_DRAM_reg/4a49645e895173d28bcd2568689200779c7b3d986a47e052f3560f409ad86fb1.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "4a71b7a70879183dd8bdce1fab8353ae66a3570042b54defc7d5b48e9b620999", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "d1e67c640bce5719fb161fdc4c001e8916c22064bc859fc333ad8c74003ba444", "generation_seconds_batch": 20.431517314980738, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.554577350616455, "target_z": -0.22070314339244307, "ce": 6.213237762451172, "genuineness": 4.519899368286133, "identity": 0.6226257085800171, "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_DRAM_reg/4a71b7a70879183dd8bdce1fab8353ae66a3570042b54defc7d5b48e9b620999.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": "4a7872f9eb16e6415f10a8e01a098a521c2b98911e33df4625645f41b8bccdd9", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "439f5f9f0d446b9d95bc1d59da9cd810bac813da89c7cfdd500095bc0a78a994", "generation_seconds_batch": 13.750861493987031, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.354781627655029, "target_z": 1.3593749836511617, "ce": 5.33954381942749, "genuineness": 4.118232727050781, "identity": 0.5305957198143005, "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_DRAM_reg/4a7872f9eb16e6415f10a8e01a098a521c2b98911e33df4625645f41b8bccdd9.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": "4b6d6daacfa07dcfa720f1267d362786af48d6d019e1450413315b078be7c224", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The road turns left behind the bridge.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "da43886ba40f8f59157cf71cddaf9ac4c95ac380b59831d8493d1471d4ad3a6a", "generation_seconds_batch": 15.32835327507928, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.586204528808594, "target_z": 1.5624998561302232, "ce": 6.09409761428833, "genuineness": 3.986435890197754, "identity": 0.4091089963912964, "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_DRAM_reg/4b6d6daacfa07dcfa720f1267d362786af48d6d019e1450413315b078be7c224.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "4c59ee96554526843acaa0b0923d1b9599cf6f4ec5099481c465c9315b315cdc", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "6bcaa12eec04cc11a0acadebfad5583e7357f82341032e2452a9e7cfb89abe91", "generation_seconds_batch": 14.563995521049947, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.509197473526001, "target_z": 0.6171875310627927, "ce": 5.919358730316162, "genuineness": 4.369274139404297, "identity": 0.3375283479690552, "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_DRAM_reg/4c59ee96554526843acaa0b0923d1b9599cf6f4ec5099481c465c9315b315cdc.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "4c859308eb64a4cc434acce75b453841d18c04d7b5f32f884ffd0dd9029b66b7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "a9cf359fea3cf3fc4bddf49422d1e4cb1ce8d6f8b7ad632e4661a532f5d1f296", "generation_seconds_batch": 20.50148458406329, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9631426334381104, "target_z": 1.015624990190697, "ce": 4.737886428833008, "genuineness": 4.068024158477783, "identity": 0.6825704574584961, "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": 0.03999999999999915, "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_DRAM_reg/4c859308eb64a4cc434acce75b453841d18c04d7b5f32f884ffd0dd9029b66b7.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": "4d093e4e19dde703a5794ab305b69c6e81b50409266ba9ee937fc5cdcfabf186", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I wrote the number on the envelope.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "50b00de5be6553dc504b9e2a97b11c465f60b7d46781ad1be6a90c663044b0a2", "generation_seconds_batch": 15.44706453208346, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.123358726501465, "target_z": 1.1562501111721002, "ce": 5.824047088623047, "genuineness": 3.873466968536377, "identity": 0.3750186860561371, "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_DRAM_reg/4d093e4e19dde703a5794ab305b69c6e81b50409266ba9ee937fc5cdcfabf186.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "4d3b48a5ec74174184bcf0c0febfabfe490f32cff7b6580bcc059408a2f28c8a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "3600200112774fbf3496bc7b56abad8d9e523c89fc133c14888c023bf8f4b66b", "generation_seconds_batch": 20.619994111009873, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8382930755615234, "target_z": 0.028320268562497135, "ce": 6.125868320465088, "genuineness": 4.055472373962402, "identity": 0.44050562381744385, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/4d3b48a5ec74174184bcf0c0febfabfe490f32cff7b6580bcc059408a2f28c8a.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "4e09047428fd6098de0f689a40d2dc8b6d1d753e25722f2cc476a2573bf3ce99", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "afc036df5ab22bbb4c745f076632958efd228b51b280b9f63edaf6451a6c7a80", "generation_seconds_batch": 21.018774896045215, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9167320728302, "target_z": 0.09716786851118532, "ce": 5.959072113037109, "genuineness": 4.080576419830322, "identity": 0.5788040161132812, "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_DRAM_reg/4e09047428fd6098de0f689a40d2dc8b6d1d753e25722f2cc476a2573bf3ce99.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "4e3144a35d10fc845cf985587ac32092efb68d95c661298b161cc646c488a48b", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "445ecd3c89945478d13e18af1f636112741cf1c6e322bae753e61bdf1f5fd425", "generation_seconds_batch": 21.718066281056963, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.58040452003479, "target_z": 0.6796874918255809, "ce": 5.4666266441345215, "genuineness": 4.268857479095459, "identity": 0.23705920577049255, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo. Sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/4e3144a35d10fc845cf985587ac32092efb68d95c661298b161cc646c488a48b.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "4f529609d1c5854caf7159743841a81326f54233a2d14b50afbb5fcd0b8c1264", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "0d5108a5152fa033d8d922fc388290969e082784f44ed5faaed4753243410ec7", "generation_seconds_batch": 16.488415655097924, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3289544582366943, "target_z": 0.4589843484331378, "ce": 6.23706579208374, "genuineness": 3.8671910762786865, "identity": 0.3253951370716095, "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_DRAM_reg/4f529609d1c5854caf7159743841a81326f54233a2d14b50afbb5fcd0b8c1264.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "4f6aa1d843f5d79400ee10d995fc5ced22625aca1c8c35920c3f760a86cde8d6", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "d68432f3d40f872a64a8f2bdf375af46d6fae5e2c3ffdfdfa3b1d9a56d19a21e", "generation_seconds_batch": 14.919987099943683, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4201884269714355, "target_z": 0.5390623708441776, "ce": 6.268836498260498, "genuineness": 4.400654315948486, "identity": 0.5742249488830566, "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_DRAM_reg/4f6aa1d843f5d79400ee10d995fc5ced22625aca1c8c35920c3f760a86cde8d6.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "4fd31075d7dbd7f06c35eb21f67beabc181d482be597f06a303ee5a80e1a6866", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "76d9ae81b5c582df0ff9d918ddbadc4f736a75a30183b1447b66baaadb4f5503", "generation_seconds_batch": 20.86922641703859, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.061052322387695, "target_z": 1.1015624362395307, "ce": 5.708878040313721, "genuineness": 4.287685871124268, "identity": 0.8957575559616089, "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": 0.03999999999999915, "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_DRAM_reg/4fd31075d7dbd7f06c35eb21f67beabc181d482be597f06a303ee5a80e1a6866.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": "4fea7d583a200bcfb8d74c74f039b755708418d06a16595783a6900e4ed1777c", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "e2acbda5236dfb95e973e51c093d5765009fa452e38a0ff94d8b14ad191b57ee", "generation_seconds_batch": 15.714376059942879, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0307748317718506, "target_z": 0.1972656581063975, "ce": 5.871703147888184, "genuineness": 4.557555198669434, "identity": 0.7612131237983704, "hypothesis": "Quelqu'un a déplacé la petite chaise bleue, elle était ici, près de la grande bibliothèque. Nous la remettrons après le désonné.", "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_DRAM_reg/4fea7d583a200bcfb8d74c74f039b755708418d06a16595783a6900e4ed1777c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "504fa4d9799fabad90c3b6a07cc2bc2ab28566518cb65553675c5220a47a1d59", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "9b9bac0ad5664de9702baca68807225514eebafdde4bdeb162c6c00674aaeeb8", "generation_seconds_batch": 20.81285622192081, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.705016851425171, "target_z": 0.7890624231604602, "ce": 6.070269584655762, "genuineness": 4.030367851257324, "identity": 0.16721801459789276, "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_DRAM_reg/504fa4d9799fabad90c3b6a07cc2bc2ab28566518cb65553675c5220a47a1d59.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "50d46993c6359c7f79261e74a27e3e2a052cdffe0d55413c73c1a140493c4bf8", "bucket": "vn_S_DRAM_reg", "phase": "wording", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "67a095890d2579d2b4929765dcea86bef82174e75ea87b5f68a786d8531399b1", "generation_seconds_batch": 16.879051913972944, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.409937858581543, "target_z": -0.3476563456407039, "ce": 6.23706579208374, "genuineness": 4.68307638168335, "identity": 0.5955392718315125, "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_DRAM_reg/50d46993c6359c7f79261e74a27e3e2a052cdffe0d55413c73c1a140493c4bf8.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "50eea355e2c0ac2ae16976962dadc31eb3dbe74b27dcda2916d513cc8dc06410", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "5191036d009f170a5af7425d0989092ffe2c0627f516662328d677ef64cbc4d6", "generation_seconds_batch": 15.824774271110073, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5893054008483887, "target_z": 0.6874999869209294, "ce": 6.157639026641846, "genuineness": 4.331617832183838, "identity": 0.7569443583488464, "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_DRAM_reg/50eea355e2c0ac2ae16976962dadc31eb3dbe74b27dcda2916d513cc8dc06410.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "50f2de1e2867a8f724ecd05be34ced0d7690a3afb17bcc8592366b703eda0b56", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "3200ee9031b98479be622522dcb27b6f50736fb566fc5ac2dd3b0a6b2fa0e101", "generation_seconds_batch": 24.770493646035902, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8563320636749268, "target_z": 0.9218750490465148, "ce": 5.689021587371826, "genuineness": 4.369274139404297, "identity": 0.6940009593963623, "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_DRAM_reg/50f2de1e2867a8f724ecd05be34ced0d7690a3afb17bcc8592366b703eda0b56.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "51c27045bc8ce5dd9995da4469702d7895c49a4857f6a092dedef95ffe257443", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "2234cfc43cb64f6d6e976159c1044a3bf312c4c412bda9eb2f00d774050ce1e1", "generation_seconds_batch": 19.869309452013113, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4468913078308105, "target_z": 0.5625000653953531, "ce": 6.308549880981445, "genuineness": 4.607763767242432, "identity": 0.4741944670677185, "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_DRAM_reg/51c27045bc8ce5dd9995da4469702d7895c49a4857f6a092dedef95ffe257443.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "5202e84ead80cdd0c65fd5bd591570dbe945c5b3fa9ede00c00d7aeb42e72c35", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "bbae29e49b7642f375f563e8e81bb168b41f3e2c8b60d3048ceda7004c1a6b4b", "generation_seconds_batch": 21.611957262037322, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.408186912536621, "target_z": 1.406249954223253, "ce": 5.474569320678711, "genuineness": 4.18099308013916, "identity": 0.761627733707428, "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_DRAM_reg/5202e84ead80cdd0c65fd5bd591570dbe945c5b3fa9ede00c00d7aeb42e72c35.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "5303a2303c5ceb978d270ad283469a80088271a7b978cec6ea27852ad5ba762b", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "10c952e2e476989de040445dd57fb2fdeb01db209412aed6fc8a3962d2d7fd8e", "generation_seconds_batch": 21.549519784050062, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7094674110412598, "target_z": 0.7929687753406993, "ce": 6.117925643920898, "genuineness": 4.168440818786621, "identity": 0.7103508710861206, "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_DRAM_reg/5303a2303c5ceb978d270ad283469a80088271a7b978cec6ea27852ad5ba762b.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "53511dabce9d017200d3112edacc69ff50541595c16b3977b4aaf9a108a5f333", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "e77bdf3491ffd443e48908c608d53bf75b36eaea48750b4a1597ded9bbaef78b", "generation_seconds_batch": 22.47039658599533, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3231544494628906, "target_z": -0.42382801587150454, "ce": 6.1020402908325195, "genuineness": 4.400654315948486, "identity": 0.5223707556724548, "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_DRAM_reg/53511dabce9d017200d3112edacc69ff50541595c16b3977b4aaf9a108a5f333.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "536e51a9a1b9d07a843445e5423730e472cd222ea0ea2f5ef5d378ce0bb9e486", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "114d1cf3bb4837af3e4592b9dbe3cb0b2e0e40bc62078baee4df815022e837b4", "generation_seconds_batch": 20.835940978024155, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.782024383544922, "target_z": 1.7343751667581504, "ce": 5.4944257736206055, "genuineness": 3.911123275756836, "identity": 0.3620527982711792, "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_DRAM_reg/536e51a9a1b9d07a843445e5423730e472cd222ea0ea2f5ef5d378ce0bb9e486.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "544d758cb0e02cd00dc354fd5178bb941a15e75688708d7c2e0d5e6828e87fd1", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "85f1ece6871c560656159dd58c5779c92e0c529935e2d27363c7ae74f83e703d", "generation_seconds_batch": 21.198350839084014, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.923088550567627, "target_z": 0.9804686576290638, "ce": 6.046441555023193, "genuineness": 4.256305694580078, "identity": 0.44612348079681396, "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_DRAM_reg/544d758cb0e02cd00dc354fd5178bb941a15e75688708d7c2e0d5e6828e87fd1.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "dd5261319a7ca4358ce2884dd40007ac339087a9ca5572b24c70868b657a8360", "generation_seconds_batch": 14.901561959064566, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9782450199127197, "target_z": -0.726562462397672, "ce": 6.181467056274414, "genuineness": 4.425758361816406, "identity": 0.6583829522132874, "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_DRAM_reg/546203f143d2dbd908be3097fd76d41d76e93afd1487fa15c64a3d119dff5d49.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "54ae23efa011b8a7c0e40193d816cf1168813a4957990d20bfe9dfeac812eec3", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "d43a192f2b03d96564929d5b177755c2f466c26ae7e8ac52296f51eee4b5ce8e", "generation_seconds_batch": 21.347048301016912, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.095306158065796, "target_z": 0.2539061952313918, "ce": 6.046441555023193, "genuineness": 4.293961524963379, "identity": 0.5243228673934937, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/54ae23efa011b8a7c0e40193d816cf1168813a4957990d20bfe9dfeac812eec3.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "54b2f879f8dfcee1496e7ea786035a57df2e283a182b34c02bc7d5ca2aabc2b7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "c655775364902b7d5ec42b161db149fab6bd8df794fa777751ac3cad97465e8a", "generation_seconds_batch": 20.80017586098984, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1798646450042725, "target_z": 0.3281250032697677, "ce": 5.935244083404541, "genuineness": 4.632867813110352, "identity": 0.797407865524292, "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_DRAM_reg/54b2f879f8dfcee1496e7ea786035a57df2e283a182b34c02bc7d5ca2aabc2b7.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "54c2acfb8afcde2e05e808e5dcea31a799b81bfe4cab9c853246962531dd1a51", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "c3b1cf31e11174f029c838a725b53458afca9c5bcc118ec3fdf5ce767e86a27a", "generation_seconds_batch": 20.445291862939484, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.364557981491089, "target_z": 0.4902343288145319, "ce": 6.062326908111572, "genuineness": 4.432034492492676, "identity": 0.5700452327728271, "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_DRAM_reg/54c2acfb8afcde2e05e808e5dcea31a799b81bfe4cab9c853246962531dd1a51.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "5523920d56ec4b1d7e02da719c27d85f28262e304d5275960e7576d31ac3a8bc", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "81dd1ccd1f5b180be70e88f13839822f9e2525d39a8e973528dfca9e3a7546b2", "generation_seconds_batch": 21.611957262037322, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.176764011383057, "target_z": 1.2031250817441914, "ce": 5.708878040313721, "genuineness": 4.218649387359619, "identity": 0.7408483028411865, "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_DRAM_reg/5523920d56ec4b1d7e02da719c27d85f28262e304d5275960e7576d31ac3a8bc.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "55de472f4bb70e8f384162015592fb7eac79120207448081f10ce2197fc3035e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "0bc0eaf7cfcc521c9226d4d0610cc26eaad335df5b0666296d90ac4a5e00c77a", "generation_seconds_batch": 20.729799936059862, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5982062816619873, "target_z": 0.6953124820162779, "ce": 5.009922981262207, "genuineness": 4.293961524963379, "identity": 0.16924472153186798, "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_DRAM_reg/55de472f4bb70e8f384162015592fb7eac79120207448081f10ce2197fc3035e.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "563b4562daa259e9017842e6a7b9bd58532692ffee12b1f0798b52d50ae48c91", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "9466e1b6080854cbd4ccabf3a9c30e630ebbe5bc8484a99161122dd61b27f822", "generation_seconds_batch": 20.22721594106406, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.674739360809326, "target_z": -0.11523435497267312, "ce": 6.340320587158203, "genuineness": 4.507347106933594, "identity": 0.44443345069885254, "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_DRAM_reg/563b4562daa259e9017842e6a7b9bd58532692ffee12b1f0798b52d50ae48c91.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "5696eb6a8572cc27ac50a40036352a44d772ed4d53c5ef742513013d004a5b4c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "bd00e1aa9a8844677932cf5f76236ad4b49eb53a137d6ac7d3688589a48a176d", "generation_seconds_batch": 21.26746656699106, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4735939502716064, "target_z": 0.5859375506813986, "ce": 6.292664527893066, "genuineness": 4.507347106933594, "identity": 0.5346795320510864, "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_DRAM_reg/5696eb6a8572cc27ac50a40036352a44d772ed4d53c5ef742513013d004a5b4c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "56b317318133d7d78c1b693dc8d0de76a7e7d9e66dd1b2f669558dde409b0b5a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "d6ba4f0defc0239008af9ce23c993611a19bf6ab905edc901ca0dbc2fb704938", "generation_seconds_batch": 21.890862816013396, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9097373485565186, "target_z": 0.968750019618606, "ce": 6.1894097328186035, "genuineness": 4.13078498840332, "identity": 0.17289596796035767, "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_DRAM_reg/56b317318133d7d78c1b693dc8d0de76a7e7d9e66dd1b2f669558dde409b0b5a.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "574d1c79cd9b22633fc0590c91cf4ae3d9f57dc468a0dea149af7451a2efa1e4", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "225ac814a26717ffce08441b9fdd998da94a7a2293dceafc5ce790bbd6a044ec", "generation_seconds_batch": 22.47681180201471, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.750953197479248, "target_z": -0.048339826481539575, "ce": 6.284721851348877, "genuineness": 4.400654315948486, "identity": 0.6096036434173584, "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_DRAM_reg/574d1c79cd9b22633fc0590c91cf4ae3d9f57dc468a0dea149af7451a2efa1e4.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "578ec3c949f3e7992b8682f77736519a2e5c6374b94c241b759683b2b3b02c65", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "369d51b061d62041be7b10f412c7d93512784a1c30ba0979956f8c92e0560f13", "generation_seconds_batch": 20.773486281977966, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.105556964874268, "target_z": 1.1406251209814033, "ce": 5.085378646850586, "genuineness": 4.281409740447998, "identity": 0.15569038689136505, "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_DRAM_reg/578ec3c949f3e7992b8682f77736519a2e5c6374b94c241b759683b2b3b02c65.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "580bf7ea8fd52f3b30485a7f3e3cfe5998ccf98145c6f5f207c656ad6d91b8eb", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "309096f2ea66e57d59a75b2f9848ddbb894b7416ec972d4f6ad3036eb45d0e8d", "generation_seconds_batch": 20.50956071401015, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7420523166656494, "target_z": -0.05615232157688809, "ce": 6.292664527893066, "genuineness": 4.457139015197754, "identity": 0.6446672081947327, "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": 0.03999999999999915, "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_DRAM_reg/580bf7ea8fd52f3b30485a7f3e3cfe5998ccf98145c6f5f207c656ad6d91b8eb.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "5953ba6c39a0f4dcd8129314bd85aa0b85696de764dd7eb60e1861142a6961f7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "a4b7bb1e8b658e396cc93265d3d2cc6b7a882b778ffd7ab4b5ee496a83bfd3de", "generation_seconds_batch": 22.04210061696358, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.629359483718872, "target_z": 0.7226563194825627, "ce": 5.673136234283447, "genuineness": 4.281409740447998, "identity": 0.2523745894432068, "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": 0.03999999999999915, "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_DRAM_reg/5953ba6c39a0f4dcd8129314bd85aa0b85696de764dd7eb60e1861142a6961f7.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "5957ec108d0c59250846a052f9ef65d53ab3416a5508e8187132b12c033374b7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "eaf0425a8de94af92b3daab8c6cbede2d305ad4fe5791ce2a83d265f9674921e", "generation_seconds_batch": 20.758678595069796, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.926189422607422, "target_z": 0.10546878841976995, "ce": 6.308549880981445, "genuineness": 4.080576419830322, "identity": 0.3700297772884369, "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_DRAM_reg/5957ec108d0c59250846a052f9ef65d53ab3416a5508e8187132b12c033374b7.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "599e80eb24933bf463ecbd209d562bc7bc3abcb1445990b34157c3779d0a2cb0", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "50fa9fafe51c7c6199eaa758f0b29bcbf72b5fff1fda44e3b5434f973d29efdd", "generation_seconds_batch": 19.956902265083045, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8385300636291504, "target_z": 0.9062498495906879, "ce": 6.268836498260498, "genuineness": 4.557555198669434, "identity": 0.5842000246047974, "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_DRAM_reg/599e80eb24933bf463ecbd209d562bc7bc3abcb1445990b34157c3779d0a2cb0.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "59d3d4949bbfed61ed04bea18599cd7bb09a322d29e966ac1357db32abd15409", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "18cce6cbd652db524dfa93215c15fba9721656bdb76fbe006280b66fb68dcf73", "generation_seconds_batch": 20.627072542905807, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.317828416824341, "target_z": 0.4492187818802346, "ce": 6.197352409362793, "genuineness": 4.457139015197754, "identity": 0.3778488337993622, "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_DRAM_reg/59d3d4949bbfed61ed04bea18599cd7bb09a322d29e966ac1357db32abd15409.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "59d817e173837055c2628e1b5fd8641ba40ca3568f96b5350e427df8bd95f2ff", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "7dd274d42f20ae471398a0849f394b269fea6b3301ab1a382f13c805fb0bb2e6", "generation_seconds_batch": 21.94769762188662, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.319178104400635, "target_z": 1.3281250032697676, "ce": 5.315715789794922, "genuineness": 4.3127899169921875, "identity": 0.5327935218811035, "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_DRAM_reg/59d817e173837055c2628e1b5fd8641ba40ca3568f96b5350e427df8bd95f2ff.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "59e05d9f9532f1ea654ba8852e02b9a9d522922d181a10db0fa6f42842e9b08c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "5b7c00312ca7246adb795573b50ba42947102b7f5ec7672a326143721b570e3f", "generation_seconds_batch": 19.5041069079889, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2497220039367676, "target_z": -0.48828125735697725, "ce": 6.09409761428833, "genuineness": 4.325342178344727, "identity": 0.5100222826004028, "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_DRAM_reg/59e05d9f9532f1ea654ba8852e02b9a9d522922d181a10db0fa6f42842e9b08c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "59ea2863d869d0706d924ef98d5196338ef51cc5332d8cad3ff0eaa4f5ec23ef", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "82c9c19f04ad06949f3fed7ca9263d548de52346c25ce1b0bb3e5d37f2819e7b", "generation_seconds_batch": 21.247972128912807, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.90083646774292, "target_z": 0.9609375245232574, "ce": 5.272031307220459, "genuineness": 4.105680465698242, "identity": 0.6234188675880432, "hypothesis": "J'ai trouvé la petite boîte, elle était près d'eux, la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "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_DRAM_reg/59ea2863d869d0706d924ef98d5196338ef51cc5332d8cad3ff0eaa4f5ec23ef.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "5a3027756f88854183272f44b2af440cfd811b12c7af26f5200a9f0b8c038a67", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "c1c36cb220ed46f74027b20dc5e277a686444c0303536a6bfad26e9ba0cd3e5d", "generation_seconds_batch": 21.95685715507716, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.606870174407959, "target_z": -0.17480460391656433, "ce": 6.324435234069824, "genuineness": 4.118232727050781, "identity": 0.532127857208252, "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_DRAM_reg/5a3027756f88854183272f44b2af440cfd811b12c7af26f5200a9f0b8c038a67.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "5a746734a18e164f572da1ad66676a6263a6f5e0d42927b92c2cc282f45f438a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "a4425becae7656e024aaaa316d95475e1a6bd7ce763c0978aaa2603551b6c027", "generation_seconds_batch": 19.98608150705695, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8719496726989746, "target_z": 0.057861389892954605, "ce": 5.967014789581299, "genuineness": 4.325342178344727, "identity": 0.6872197389602661, "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": 0.03999999999999915, "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_DRAM_reg/5a746734a18e164f572da1ad66676a6263a6f5e0d42927b92c2cc282f45f438a.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": "5ac955e4fdc933182016664ba0b9c268a44d88129410b2d7c81b69767ed731bd", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "143f4125fff5973ade54dc33a18442a4e89aff56829034f60516ec6d963bea36", "generation_seconds_batch": 21.12764885101933, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3801345825195312, "target_z": 0.5039062475476742, "ce": 6.046441555023193, "genuineness": 4.18099308013916, "identity": 0.5397111177444458, "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_DRAM_reg/5ac955e4fdc933182016664ba0b9c268a44d88129410b2d7c81b69767ed731bd.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "5acea1ee34888a2abaee40d0a9a3cc0293f6cf1fc52258c1e16c4f9e7c0e7f48", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "7fcda88fc68b44e27dd97622cdcb374c8579ee239c4f6c846b3606d3a4bcf484", "generation_seconds_batch": 21.926073814043775, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2555222511291504, "target_z": 0.394531316212795, "ce": 6.23706579208374, "genuineness": 4.017816066741943, "identity": 0.24819493293762207, "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_DRAM_reg/5acea1ee34888a2abaee40d0a9a3cc0293f6cf1fc52258c1e16c4f9e7c0e7f48.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "5ae1c35a68f1e449af6b5ec9a5fedf174ae8259a1111c770d216ce31d354263e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "85568cd09e0ccd551648e3fb530f6d8c1c40548d53f583d503f9e4b7e8ca3a48", "generation_seconds_batch": 21.923580467002466, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.678077220916748, "target_z": -0.11230464315377618, "ce": 6.2767791748046875, "genuineness": 4.545003414154053, "identity": 0.5708436965942383, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist. Ich.", "wer": 0.05263157894736842, "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_DRAM_reg/5ae1c35a68f1e449af6b5ec9a5fedf174ae8259a1111c770d216ce31d354263e.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "5b7ff4da1e74a0b88aabaa002e67afb5fdd84594f3db89377dac17884f726221", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "9dd484e8758fcfd4ad08ea86c40d6f5816c5a232b6b0a2a8b4dad6505671faf3", "generation_seconds_batch": 20.811232559964992, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.723694324493408, "target_z": -0.0722655273156913, "ce": 5.744620323181152, "genuineness": 4.281409740447998, "identity": 0.5373595952987671, "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_DRAM_reg/5b7ff4da1e74a0b88aabaa002e67afb5fdd84594f3db89377dac17884f726221.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "5b99286f6b513fcfa4a64a9ba1e35086b9b0e3e53ecbbfdd6442e7768fc6deb4", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "f2f48b96e1064e67ec12445fd8f18143c0945aeceee8748cddb3d00b2701842c", "generation_seconds_batch": 21.033383051049896, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8874850273132324, "target_z": 0.9492186772476697, "ce": 5.9431867599487305, "genuineness": 4.118232727050781, "identity": 0.36715441942214966, "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_DRAM_reg/5b99286f6b513fcfa4a64a9ba1e35086b9b0e3e53ecbbfdd6442e7768fc6deb4.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "5bad30431980a882e987b765a205e9d5ba338adc8e54e9b23fada438e5d04af4", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "980a1ae7a07a3f33c45c6ee09609bada1f7a714cbf429c8b28553793de2531f1", "generation_seconds_batch": 20.28390231996309, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.193215847015381, "target_z": 0.3398436412802255, "ce": 6.054384231567383, "genuineness": 4.143336772918701, "identity": 0.11347702145576477, "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_DRAM_reg/5bad30431980a882e987b765a205e9d5ba338adc8e54e9b23fada438e5d04af4.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "5bb9ba856a4012d2b6d4b7ef06b8fc1a0b6397e563de8d365eee3a3d977d6bfa", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "d1730dd4ea69dc21ebc54edf2ca54de4c635fda42e2ced6b0499e375d91683bb", "generation_seconds_batch": 20.80735570890829, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.514997482299805, "target_z": 1.499999895367435, "ce": 6.070269584655762, "genuineness": 4.080576419830322, "identity": 0.3089814782142639, "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_DRAM_reg/5bb9ba856a4012d2b6d4b7ef06b8fc1a0b6397e563de8d365eee3a3d977d6bfa.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": "5bd7d8a18c6dda28cceb75ef3d45fd22e74b9201a3d03a5cd86ebc62c9fa29ef", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "db255ffa7cc948a214927c8a7913a2d445445712a0c217105e4b6919cc70c65c", "generation_seconds_batch": 20.923600922105834, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.749521255493164, "target_z": 0.8281248986372027, "ce": 5.7684478759765625, "genuineness": 4.268857479095459, "identity": 0.42121589183807373, "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_DRAM_reg/5bd7d8a18c6dda28cceb75ef3d45fd22e74b9201a3d03a5cd86ebc62c9fa29ef.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": "5bdc422626b37b36b5d32a7ceb56ab9cdb01706c7496b54e8b22d0de1433adbe", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "2075d4f67edb4ef67a17242698414a3280cf2a637488f9466819841699ea9962", "generation_seconds_batch": 15.638461338938214, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6961159706115723, "target_z": 0.7812499280651116, "ce": 6.062326908111572, "genuineness": 4.425758361816406, "identity": 0.43534159660339355, "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_DRAM_reg/5bdc422626b37b36b5d32a7ceb56ab9cdb01706c7496b54e8b22d0de1433adbe.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "5c3f3c8541fa9cbc0fce9b3dfbec57f79eb88c2d9d88907ec861c20147fcd5e2", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "3a12c8262483b74e838755e5f5e14207417b69395188c4a7c286597d4dbdd33e", "generation_seconds_batch": 22.153786478098482, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.927539110183716, "target_z": 0.9843750098093029, "ce": 5.470598220825195, "genuineness": 4.256305694580078, "identity": 0.7052253484725952, "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_DRAM_reg/5c3f3c8541fa9cbc0fce9b3dfbec57f79eb88c2d9d88907ec861c20147fcd5e2.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "5c9542ecdea81825276b9d7745838808be5b8baf9d668d7aee93c9513eed3c4b", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "3afdbec8c9f627f6b2fa6daf2a223f9d4765c5fa6d6aa119ea3f9b1d901410bb", "generation_seconds_batch": 21.181547805899754, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7940258979797363, "target_z": 0.8671875833790752, "ce": 5.109206199645996, "genuineness": 4.432034492492676, "identity": 0.572957456111908, "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_DRAM_reg/5c9542ecdea81825276b9d7745838808be5b8baf9d668d7aee93c9513eed3c4b.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": "5ccf9f671623bf3933183c83b77560caabb80ad10853eaea8e71063402064888", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "585dfc3cb1b8b168d48da706a06bd5b1984f18e8e083fa165348b16078382c3a", "generation_seconds_batch": 15.934285057010129, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.310276985168457, "target_z": 1.3203122989092893, "ce": 5.5539960861206055, "genuineness": 4.093128204345703, "identity": 0.4101783335208893, "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_DRAM_reg/5ccf9f671623bf3933183c83b77560caabb80ad10853eaea8e71063402064888.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": "5db1625e6efcd46e5018b9eafa8f0c93265fd6744bebbf5ca8c8c8c941ef7c83", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "4f74184f641ca14babcebdcbd702fced726d169b00e59be1bc698972ae40a145", "generation_seconds_batch": 14.225342463003471, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7673230171203613, "target_z": 0.8437498888278997, "ce": 5.736677646636963, "genuineness": 3.911123275756836, "identity": 0.30821049213409424, "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_DRAM_reg/5db1625e6efcd46e5018b9eafa8f0c93265fd6744bebbf5ca8c8c8c941ef7c83.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "5dff7a980fd520161221f1b5327a1bc6cba02cb55aa91abc7e55eac830eb0d31", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "1c07206e10c856025ba1f3ee3e556449cfecd0731dc7d0b08ae37a1c107221e3", "generation_seconds_batch": 16.24833213200327, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6958789825439453, "target_z": -0.09667965296307915, "ce": 6.149696350097656, "genuineness": 4.394378662109375, "identity": 0.7094594836235046, "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_DRAM_reg/5dff7a980fd520161221f1b5327a1bc6cba02cb55aa91abc7e55eac830eb0d31.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "f222a2159688fc4302f16f9433948f4fe70be47b0351b68549adfae823ba1c4a", "generation_seconds_batch": 15.261628977954388, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.610208034515381, "target_z": -0.17187489209766738, "ce": 6.133810997009277, "genuineness": 4.381826400756836, "identity": 0.7869293689727783, "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_DRAM_reg/5e0af124ad7421917e897f2876d49b25e213a1bc87e01fe43f09b9c255c2a32f.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "5ee5c0382d11668c76dfd83e09660e0679013820a6cb9a898a315d78995a6413", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "108feb393caf4c55861b0b4033952ed64db268674457d3e922d6d3dfcc516fd4", "generation_seconds_batch": 14.900078602018766, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.123358726501465, "target_z": 1.1562501111721002, "ce": 5.418970584869385, "genuineness": 4.400654315948486, "identity": 0.7791836261749268, "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_DRAM_reg/5ee5c0382d11668c76dfd83e09660e0679013820a6cb9a898a315d78995a6413.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "5f776de5b4d94b8381a9131ea90054a7fb3218d7248f075d7404b6fa70ac0b6d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "752e7e696b57ea87d943e29d2281c58922414cfc9a340a48ddd5ca08f7f72ea2", "generation_seconds_batch": 22.279276649001986, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.016548156738281, "target_z": 1.062500170027918, "ce": 5.95112943649292, "genuineness": 4.024092197418213, "identity": 0.5991030931472778, "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_DRAM_reg/5f776de5b4d94b8381a9131ea90054a7fb3218d7248f075d7404b6fa70ac0b6d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "5f905c75e65b5a9aeba3dfbe22e22117e715571511db6622c64d4972fbc87538", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "9c1162e487d7749fb0468ffb15c595179d0ca5e6ef93702c71976f974e62e801", "generation_seconds_batch": 21.82172237301711, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.918638229370117, "target_z": 0.9765625147139545, "ce": 6.062326908111572, "genuineness": 4.419482707977295, "identity": 0.35104846954345703, "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_DRAM_reg/5f905c75e65b5a9aeba3dfbe22e22117e715571511db6622c64d4972fbc87538.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "5f93b1b712042f7c138bc372934907c491a01abbbd225bb3e33c686e4f44fd6d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "5908b4a612eb22d4775f86dfe57c8d1a79a424261f1fbe7dce115062f40ca3a8", "generation_seconds_batch": 20.839448106009513, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.357882499694824, "target_z": 0.4843751144418679, "ce": 5.7684478759765625, "genuineness": 4.482243061065674, "identity": 0.7349766492843628, "hypothesis": "J'ai trouvé la petite boîte elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'est ce qu'elle contient.", "wer": 0.10526315789473684, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_DRAM_reg/5f93b1b712042f7c138bc372934907c491a01abbbd225bb3e33c686e4f44fd6d.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": "5faffc5a5c3cfd0256355bf89d1fee5d906fe6434a0637f20d74bb1aa4336fe1", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "eec1c3704157152437fe4d4f7abb2560584bc155603c1c3f18a3c6fc5859bf5f", "generation_seconds_batch": 20.857154129073024, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6605124473571777, "target_z": 0.7499999476837175, "ce": 6.300607204437256, "genuineness": 4.268857479095459, "identity": 0.34090059995651245, "hypothesis": "De weg gaat linksaf naar de brug, loopt verder, tot aan het kleine houten hek. Ik wacht aan de andere kant.", "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_DRAM_reg/5faffc5a5c3cfd0256355bf89d1fee5d906fe6434a0637f20d74bb1aa4336fe1.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "5fe987aea94fa53a80ef36ca5aca2b70ef5f643d9044e74f204fd80407871df2", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "a57a415b2ba5808b41367273a716b6a1bea71e0451e32148d612935522634215", "generation_seconds_batch": 21.32284016907215, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.3280792236328125, "target_z": 1.335937707630246, "ce": 5.411027908325195, "genuineness": 4.055472373962402, "identity": 0.499783456325531, "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_DRAM_reg/5fe987aea94fa53a80ef36ca5aca2b70ef5f643d9044e74f204fd80407871df2.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "6009eb721fa1a585441db1b1113a78d87dfc8bbeaa855d06dfea4639c041fed9", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "56594b3584568c61758e4fc8fec1aaa25b618b5ff12358ca8adb38d70a54a3a5", "generation_seconds_batch": 21.5447271540761, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.141160488128662, "target_z": 1.1718751013627973, "ce": 5.470598220825195, "genuineness": 4.137060642242432, "identity": 0.4722788333892822, "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_DRAM_reg/6009eb721fa1a585441db1b1113a78d87dfc8bbeaa855d06dfea4639c041fed9.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "600ba0fb9c57142f89732e761ea4cadd380468e762b27512c445e27222d0c0c7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "5322dbaa9538e42bd72d31513ad9ada6bc6a6b59a0ba1249510f37e1b1f83d31", "generation_seconds_batch": 22.961062068934552, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6853091716766357, "target_z": -0.10595700396787613, "ce": 6.197352409362793, "genuineness": 4.13078498840332, "identity": 0.6965945363044739, "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": 0.03999999999999915, "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_DRAM_reg/600ba0fb9c57142f89732e761ea4cadd380468e762b27512c445e27222d0c0c7.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "607bee53c2c538521769b4764c7e00da6caedf08027d1d006abf778a54486cef", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "532813eecfd554b1b23aac7e155aa54976e08e8b1f6b75aec08a72a8253d1521", "generation_seconds_batch": 20.858004731009714, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8211867809295654, "target_z": 0.01330570475578027, "ce": 6.157639026641846, "genuineness": 4.149612903594971, "identity": 0.5421937704086304, "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_DRAM_reg/607bee53c2c538521769b4764c7e00da6caedf08027d1d006abf778a54486cef.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "6157ec748893c312081fec17efa5129549697a38eb5d264ecbf0a92a72860e41", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "45e7c7e4c3febd19b7826e73ca65e12ca0f2089d0fbae22fd41b3d79c061433d", "generation_seconds_batch": 20.66261321795173, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.651611566543579, "target_z": 0.742187452588369, "ce": 6.173524379730225, "genuineness": 4.168440818786621, "identity": 0.11299306154251099, "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_DRAM_reg/6157ec748893c312081fec17efa5129549697a38eb5d264ecbf0a92a72860e41.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "616da6dbe86afe5c015bea7d626fed6716aff42954e0a238d13cf453c9a2dc71", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "c7a1f2153750670cb89af47168990ce90d8d6f002f3424022eb4379dfe41c883", "generation_seconds_batch": 21.857483603991568, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 5.049050807952881, "target_z": 1.9687500196186059, "ce": 5.927301406860352, "genuineness": 3.7918784618377686, "identity": 0.39213287830352783, "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_DRAM_reg/616da6dbe86afe5c015bea7d626fed6716aff42954e0a238d13cf453c9a2dc71.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "61b7f2b4d1608bcea70ad8455d6fb26486dbe2f78b62c21023a4ebd61948957f", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "d4495261ea54ffe4f1494738ac8918791be88e801ba7b8715d8842734244f82c", "generation_seconds_batch": 16.094509107992053, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.799351692199707, "target_z": -0.005859423637793869, "ce": 6.23706579208374, "genuineness": 4.256305694580078, "identity": 0.26161837577819824, "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/manual-comparison-v1/audio/vn_S_DRAM_reg/61b7f2b4d1608bcea70ad8455d6fb26486dbe2f78b62c21023a4ebd61948957f.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "621f35017015955640cb009befb8be5fa7a5cc4fdcb54ef7ae793700cc1f08de", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "70a1c54fd3eb5979cc958ab82268d9a5a8effbdd1429f726fe1176799035a352", "generation_seconds_batch": 20.70685949001927, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5002965927124023, "target_z": 0.6093750359674442, "ce": 5.967014789581299, "genuineness": 4.519899368286133, "identity": 0.519340991973877, "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_DRAM_reg/621f35017015955640cb009befb8be5fa7a5cc4fdcb54ef7ae793700cc1f08de.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "6280b11d8047c0d8aee2b5a73dfce2438c14f4c86a6f313d55a7755ab19619da", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "639bea83f6eefafb209ff5bfa852c69d8224230a5368a20ce6c9d81089b53f20", "generation_seconds_batch": 15.839338808087632, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.851644515991211, "target_z": 0.04003911583808487, "ce": 6.2767791748046875, "genuineness": 4.519899368286133, "identity": 0.5318406820297241, "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_DRAM_reg/6280b11d8047c0d8aee2b5a73dfce2438c14f4c86a6f313d55a7755ab19619da.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "6298e8a7b9f6db3e2688e650b4c33d122fad77f097d718bf7f62c603ca6fe032", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "c1aa2b3dfe2f08ad653845c2d56b1d08784e7994cf114d90cbfad6f3bcb3c46a", "generation_seconds_batch": 19.73885453992989, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.969581127166748, "target_z": 0.14355462353517026, "ce": 5.839932441711426, "genuineness": 4.557555198669434, "identity": 0.6497605443000793, "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_DRAM_reg/6298e8a7b9f6db3e2688e650b4c33d122fad77f097d718bf7f62c603ca6fe032.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "62c9667fbe8df4b5b54ad35df35c738969344a634c8653a3f17c873a583ece52", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "0716c9c6e10e183ff392f97892aaef0c79062a28ca88f4e62c271426a7744652", "generation_seconds_batch": 16.104008362977766, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.28357458114624, "target_z": 1.2968750228883736, "ce": 5.502368927001953, "genuineness": 4.174716949462891, "identity": 0.2144985795021057, "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_DRAM_reg/62c9667fbe8df4b5b54ad35df35c738969344a634c8653a3f17c873a583ece52.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "631e4103617ee95ab43cf3222a6539ac21e46e5fdcccbd0586cd4ee153c557f9", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "96dd665d34f7dbe18d21f381c61e37b7db78b5c5007942b7d1906d76f0afbb7d", "generation_seconds_batch": 20.9530884999549, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.239069938659668, "target_z": 1.2578123381465012, "ce": 5.557967185974121, "genuineness": 4.118232727050781, "identity": 0.8866397142410278, "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_DRAM_reg/631e4103617ee95ab43cf3222a6539ac21e46e5fdcccbd0586cd4ee153c557f9.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "63f842ee10b4961e381bf98bfc138491125b533fcb4502797b3b90019f0fc5cb", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "f2b288b3f9895d8c56cc4ea9dd933e627a9702d3dd0f2d5a3d5af26f7a216f64", "generation_seconds_batch": 21.65030979795847, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0850555896759033, "target_z": -0.6328125212534897, "ce": 6.006728172302246, "genuineness": 4.331617832183838, "identity": 0.6711903214454651, "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_DRAM_reg/63f842ee10b4961e381bf98bfc138491125b533fcb4502797b3b90019f0fc5cb.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "64e3201e484b312f7f083792677abe711efc13c4d5984214b0f7c2c3db5c1589", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "6fe4caa2cbbf1b3bcc6b8b45e50dee8cbf2c85b148f86718a340c41a4b39e500", "generation_seconds_batch": 21.732971908990294, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.190990686416626, "target_z": 0.3378905698226708, "ce": 6.1894097328186035, "genuineness": 4.18099308013916, "identity": 0.6986110210418701, "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_DRAM_reg/64e3201e484b312f7f083792677abe711efc13c4d5984214b0f7c2c3db5c1589.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "657c93eb424ef1f4e7f3d4f61e3699bba2a2b4c7ffedd05726026c6e3cd1fc09", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "ecbf5c0e30f1b002493849b863ada9ddb30b62fd8ad0246b688a2a06d2b12594", "generation_seconds_batch": 21.531204871018417, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.992946147918701, "target_z": 0.16406260626744879, "ce": 6.427689552307129, "genuineness": 4.545003414154053, "identity": 0.41059011220932007, "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_DRAM_reg/657c93eb424ef1f4e7f3d4f61e3699bba2a2b4c7ffedd05726026c6e3cd1fc09.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "65a98ca5bf3389c4561287917e98e39c03c80323a550ff2e55e995e725b1cc6a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "27a9ac416ade1ab04ced5feb329a82d7767328b976a60548532b18982737f82e", "generation_seconds_batch": 20.146372135030106, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4677937030792236, "target_z": -0.2968750228883736, "ce": 6.062326908111572, "genuineness": 4.607763767242432, "identity": 0.5964944362640381, "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_DRAM_reg/65a98ca5bf3389c4561287917e98e39c03c80323a550ff2e55e995e725b1cc6a.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "65af1180aa1cb261e1cb00f08e45d5bdee19719fc8c501b78932b725829bde05", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "b3af49bb1df03d9d82be698eaef4461ae3fbe828573ac10a30c85dcd0053a309", "generation_seconds_batch": 15.591938878060319, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0374505519866943, "target_z": 0.2031250817441914, "ce": 6.403861999511719, "genuineness": 4.570107460021973, "identity": 0.4491252899169922, "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_DRAM_reg/65af1180aa1cb261e1cb00f08e45d5bdee19719fc8c501b78932b725829bde05.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": "65c4b65202214dc5a7008045be7924630b2e7836bcd2dfd18a8b735e336667aa", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "67a8af25f62d099fd7521d76a38b0a108d2995ae874bb179689f7695c23779c1", "generation_seconds_batch": 15.839278992032632, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7228188514709473, "target_z": 0.804687622616287, "ce": 5.577824115753174, "genuineness": 4.331617832183838, "identity": 0.3248882293701172, "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_DRAM_reg/65c4b65202214dc5a7008045be7924630b2e7836bcd2dfd18a8b735e336667aa.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "662abbb79ee9a15a628f652ea8736c81965e5759a5232f69d6c1af056f7052dd", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "596ad4b7d30d8aac375c751ee940147b2f261c6a79fc4b646646151dd0a29381", "generation_seconds_batch": 21.53801813791506, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4544425010681152, "target_z": -0.3085936608988314, "ce": 6.3641486167907715, "genuineness": 4.65797233581543, "identity": 0.6398798823356628, "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_DRAM_reg/662abbb79ee9a15a628f652ea8736c81965e5759a5232f69d6c1af056f7052dd.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "66311c44a45ff0ce7f280db1d81e1ddaa0a1ad6cb3ab5a94e269c65cbd60a696", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "367e5ae9d33dc2148ad74f854fb96764f6d832180679977f9afac8e6451756fa", "generation_seconds_batch": 21.87371906498447, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.409937858581543, "target_z": -0.3476563456407039, "ce": 6.165581703186035, "genuineness": 4.469690799713135, "identity": 0.41498807072639465, "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_DRAM_reg/66311c44a45ff0ce7f280db1d81e1ddaa0a1ad6cb3ab5a94e269c65cbd60a696.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "6669d2b61a3070c3d9cb27e6cf07eceed58eff93ef8b1a7e0ef0cbb4316a22ac", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "f9c016e6d86b0c411ddce4959e6f7eca28c0628af74c62219cd9af7234012a49", "generation_seconds_batch": 21.531204871018417, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3044769763946533, "target_z": 0.4374999346046469, "ce": 6.006728172302246, "genuineness": 4.369274139404297, "identity": 0.47342437505722046, "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_DRAM_reg/6669d2b61a3070c3d9cb27e6cf07eceed58eff93ef8b1a7e0ef0cbb4316a22ac.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "66cfbb21cbbe4a10422bc6397992e9d64aea583854cb92cba8fc7c24c6a9ed29", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "aa105bfd98b7165916a6d2e0dc1d9aee44d43da6b6c3b7647e00730e4da0c2ee", "generation_seconds_batch": 20.516600609989837, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.333405017852783, "target_z": 0.462890700613377, "ce": 6.181467056274414, "genuineness": 4.042920112609863, "identity": 0.2394925355911255, "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_DRAM_reg/66cfbb21cbbe4a10422bc6397992e9d64aea583854cb92cba8fc7c24c6a9ed29.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "671d3b47eb6c0e4e13011b2c2d37421318e47395941459de62d27599b6105887", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "64b6cbf9de57a15c2b40c774eabaf66121e1423fbd6bbb82d7d28cd8b8a9fedd", "generation_seconds_batch": 15.519081822014414, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.052151679992676, "target_z": 1.0937501504093121, "ce": 6.054384231567383, "genuineness": 4.168440818786621, "identity": 0.29130205512046814, "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_DRAM_reg/671d3b47eb6c0e4e13011b2c2d37421318e47395941459de62d27599b6105887.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "675b58fe9cbf3a2a8aa39513803bf6e7bdcf6a331c34af3129021940b35441cb", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "5be762555c4b9ae76b46338d09b3f162b3c76f0673d543396b639781fb5f36fa", "generation_seconds_batch": 19.48480958910659, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.425988674163818, "target_z": 1.4218749444139498, "ce": 5.327630043029785, "genuineness": 4.256305694580078, "identity": 0.6398334503173828, "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_DRAM_reg/675b58fe9cbf3a2a8aa39513803bf6e7bdcf6a331c34af3129021940b35441cb.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "67f0d3498e38330cce4c873a76955352e451323af8272cabe29bd87d436c4ff2", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "b0bdab7b2fb65bd86a3ad2cab9cfaf81b14ea421d8911254df345275114e1f35", "generation_seconds_batch": 21.837467137956992, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.776224136352539, "target_z": 0.8515625931883781, "ce": 5.665193557739258, "genuineness": 4.281409740447998, "identity": 0.19497667253017426, "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/manual-comparison-v1/audio/vn_S_DRAM_reg/67f0d3498e38330cce4c873a76955352e451323af8272cabe29bd87d436c4ff2.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "68c7028451af9195a94d1bb67e29b1f5d54c68bf0db02146318cf48019374db2", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "b32281318d397dc6a3c672092cdf1205e784292475417c3b8b38c3c7a175d70d", "generation_seconds_batch": 15.837719884933904, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3231544494628906, "target_z": -0.42382801587150454, "ce": 6.109982967376709, "genuineness": 4.65797233581543, "identity": 0.562267005443573, "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_DRAM_reg/68c7028451af9195a94d1bb67e29b1f5d54c68bf0db02146318cf48019374db2.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "68cd1212d8cca1480bfba69e5b3dded38e78d352f884262b2aa0b6ed56333fba", "bucket": "vn_S_DRAM_reg", "phase": "wording", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "ce8909b67c229850f33d11a06d5abc30404fee07bcc03519abcbae9a075f25b3", "generation_seconds_batch": 15.908091485034674, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4922711849212646, "target_z": -0.27539060905988266, "ce": 6.213237762451172, "genuineness": 4.337893962860107, "identity": 0.7460219860076904, "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_DRAM_reg/68cd1212d8cca1480bfba69e5b3dded38e78d352f884262b2aa0b6ed56333fba.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "69a0939c1c80bfdf3f776e1f622e7b149107b026427f4e1a50e2ddca5e97d3c6", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "1bc4d08ebd338c7bbe9e6206f7073bca39e9f789b2db52d2ee0cd9f4b31fc17f", "generation_seconds_batch": 19.555543410009705, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2043421268463135, "target_z": 0.34960941709825855, "ce": 6.022613525390625, "genuineness": 4.256305694580078, "identity": 0.21819528937339783, "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": 0.03999999999999915, "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_DRAM_reg/69a0939c1c80bfdf3f776e1f622e7b149107b026427f4e1a50e2ddca5e97d3c6.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "69b10100853108c2af012d426bba05f47b8f96714b81eb678117a7cae53c7768", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "dc34e28e4be3571ff3ec867d928e3e0e3b58984afd214e1031b2e38a2559b8e9", "generation_seconds_batch": 21.171578763984144, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.016548156738281, "target_z": 1.062500170027918, "ce": 5.8875885009765625, "genuineness": 4.105680465698242, "identity": 0.39059484004974365, "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_DRAM_reg/69b10100853108c2af012d426bba05f47b8f96714b81eb678117a7cae53c7768.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "6a21871638b3ce3723bdfe2622f53b63236c7bbff706620f36a968f8f5fe0c38", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "644433a217f91f6defdf822d02c4ff42813394d2bab8643dd94820a85f86daf0", "generation_seconds_batch": 15.655956560047343, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.701916217803955, "target_z": 1.6640625016348838, "ce": 6.006728172302246, "genuineness": 3.923675537109375, "identity": 0.2749027609825134, "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_DRAM_reg/6a21871638b3ce3723bdfe2622f53b63236c7bbff706620f36a968f8f5fe0c38.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "84805a8a7698379c4cdf6cd76f2813b2973341be0b095b5d1d1018e1209e367a", "generation_seconds_batch": 15.367575297947042, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5701539516448975, "target_z": -0.20703122465930068, "ce": 6.419747352600098, "genuineness": 4.268857479095459, "identity": 0.3657394051551819, "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_DRAM_reg/6a61bc20d43b3e97640d8dd45682f7055c5dbaa5aa987cdb55f8a7d7a432a2c2.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "6a78f46935b94b6ff0654d9c271f8f3a315366feb966b030022b9890f60cc924", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "55a83c294dd3e4e5ded1d31bdba13b414f3182564a5fca2005d292c64e526a47", "generation_seconds_batch": 22.756419694051147, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5136477947235107, "target_z": 0.621093673977902, "ce": 5.285930633544922, "genuineness": 4.432034492492676, "identity": 0.2802509069442749, "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_DRAM_reg/6a78f46935b94b6ff0654d9c271f8f3a315366feb966b030022b9890f60cc924.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "6abde2b0f53e545ffc03dd565987187cad04655b76e489098776a0a503fe2326", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "3e144995e42566f53824f4f8b156f4543c5a2d0a93a292f350c93b79917271e8", "generation_seconds_batch": 21.159223319031298, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.773123264312744, "target_z": 1.7265624623976719, "ce": 5.689021587371826, "genuineness": 3.873466968536377, "identity": 0.13623519241809845, "hypothesis": "De deur staat nu open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.05263157894736842, "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_DRAM_reg/6abde2b0f53e545ffc03dd565987187cad04655b76e489098776a0a503fe2326.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "6b493d873e237f2a0af3c52dd967e32d1a2d7cfd0cb6953609bcff0fe6df08e5", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "1b43a79b0360b222bd9a0eab0765096e2049a5691a1a2569c27319e3e9d9816a", "generation_seconds_batch": 21.30677841603756, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.408186912536621, "target_z": 1.406249954223253, "ce": 5.927301406860352, "genuineness": 4.055472373962402, "identity": 0.2688359022140503, "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_DRAM_reg/6b493d873e237f2a0af3c52dd967e32d1a2d7cfd0cb6953609bcff0fe6df08e5.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "e43df6fc09c14e50e4c93cf0b3afbeb92e2dd797930abacbb94311d7aa98f9d1", "generation_seconds_batch": 16.136952575063333, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.601306915283203, "target_z": -0.17968759645814583, "ce": 6.252951145172119, "genuineness": 4.381826400756836, "identity": 0.5285547375679016, "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_DRAM_reg/6b973c750e3616c70f28663c29496a0c74bea4a87dff4bad65cadc5394b99131.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "6b99eb8195f78cf87d4d6ffca73ff93c2ad68ceabe71db9a4f9e484e93102756", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "3c96fd8f863ce2a44835fc775b2d1866e4f019d4e480b3128498ec14ea15c92d", "generation_seconds_batch": 22.301188891986385, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8410747051239014, "target_z": 0.030761764833287887, "ce": 6.38003396987915, "genuineness": 4.6203155517578125, "identity": 0.47760701179504395, "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_DRAM_reg/6b99eb8195f78cf87d4d6ffca73ff93c2ad68ceabe71db9a4f9e484e93102756.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": "6ba3d6152aeeff50151b4a44a0adb5b4111f7bc0862102d422511831db36f15a", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "0679f543e71afe13d0e355f1dfcc2ba820aa44adb1dfbf150fed4aa9781adf99", "generation_seconds_batch": 15.377575946971774, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 5.066852569580078, "target_z": 1.984375009809303, "ce": 5.418970584869385, "genuineness": 3.5094566345214844, "identity": 0.35646575689315796, "hypothesis": "De sleutels leggen op de keukentafel. Ik keek in de aa en daarna, naast het raam, 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_DRAM_reg/6ba3d6152aeeff50151b4a44a0adb5b4111f7bc0862102d422511831db36f15a.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "6ba5aff28bac28a06f71d5b27fb4ce6d0d89b3ff20361b5dcb56fcc6add408fd", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "3647996e6bfa4a40d8fd42b4571b1a2e46216b32123d868404c4b5716aeaa9ff", "generation_seconds_batch": 21.256687902030535, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.335630178451538, "target_z": 0.4648437720709317, "ce": 6.149696350097656, "genuineness": 4.595211505889893, "identity": 0.5649733543395996, "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_DRAM_reg/6ba5aff28bac28a06f71d5b27fb4ce6d0d89b3ff20361b5dcb56fcc6add408fd.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "6bc4ea4460be3f95180eec6bbd50c01abe6b7c06822a422fd282d833d67e836c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "b8f5cd4bdbb53123c4b830866f5c0550ec6732b54655fb6de8b48a8901fb5c7c", "generation_seconds_batch": 21.156741929939017, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8337035179138184, "target_z": 0.02429191481151385, "ce": 6.284721851348877, "genuineness": 4.507347106933594, "identity": 0.577096700668335, "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_DRAM_reg/6bc4ea4460be3f95180eec6bbd50c01abe6b7c06822a422fd282d833d67e836c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "6be4d745df39343d5274ddf6ba38f869a0382b3ad291db3caeedbe15f368bbea", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "dc778f21b48c90e8ee4311d39e36a872d013dd23b121dcd35ab0650ead336aa9", "generation_seconds_batch": 21.597625555936247, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.141160488128662, "target_z": 1.1718751013627973, "ce": 6.0146708488464355, "genuineness": 4.482243061065674, "identity": 0.5502676963806152, "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_DRAM_reg/6be4d745df39343d5274ddf6ba38f869a0382b3ad291db3caeedbe15f368bbea.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "6c0e76da13c16d9382335b2bedd195d72eca115706ad6e119425e71cdaeb3618", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "db1535e6bc46b12520544d0c6b055e99875ae946a1faa3b65105f03e9adef841", "generation_seconds_batch": 21.95685715507716, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1829652786254883, "target_z": -0.5468750752046561, "ce": 6.300607204437256, "genuineness": 4.482243061065674, "identity": 0.6699572801589966, "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_DRAM_reg/6c0e76da13c16d9382335b2bedd195d72eca115706ad6e119425e71cdaeb3618.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "6c286a69c9e369ecb5325435e28415d10e061340e118d127b19eb8910d352645", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "df8be42268cb05d632a04f964b716ac15c07e7c0cf8314727de6cb0d82414d53", "generation_seconds_batch": 20.066212925012223, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.531449794769287, "target_z": 0.636718873433729, "ce": 6.395919322967529, "genuineness": 4.482243061065674, "identity": 0.2995090186595917, "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_DRAM_reg/6c286a69c9e369ecb5325435e28415d10e061340e118d127b19eb8910d352645.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "6cb927215af3bbe82931d760bf1d010a21cb1de9b2ab0e40bd24878905c65570", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "b35bce1cbe5e3de36d49004fc8e05ab87789afcaf4f144211b4d315ab6832e86", "generation_seconds_batch": 20.66200865409337, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2221438884735107, "target_z": 0.3652344072889556, "ce": 6.109982967376709, "genuineness": 4.482243061065674, "identity": 0.5496425032615662, "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_DRAM_reg/6cb927215af3bbe82931d760bf1d010a21cb1de9b2ab0e40bd24878905c65570.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "6d2cc2d6c02004e0477630d725d6db85a6e6531691d60dc6a9cafbc4973a8199", "bucket": "vn_S_DRAM_reg", "phase": "wording", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "9406200f38892b07d62599ccba4148b436e6b773c1c9197f02702772878e2193", "generation_seconds_batch": 18.50288703397382, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6641695499420166, "target_z": -0.12451170597747009, "ce": 6.125868320465088, "genuineness": 4.2249250411987305, "identity": 0.41548606753349304, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/6d2cc2d6c02004e0477630d725d6db85a6e6531691d60dc6a9cafbc4973a8199.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "6d7d12c2c29906d285b2ced98e2683a05a51fad73a627f05b6e81d68345d83ac", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "bf34b14ae2f4cdf362607a3d30dc4221490c4d673842b2528e4258f2b263dd32", "generation_seconds_batch": 21.788887670030817, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.659162759780884, "target_z": -0.1289062737058155, "ce": 6.38797664642334, "genuineness": 4.306513786315918, "identity": 0.6604917049407959, "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_DRAM_reg/6d7d12c2c29906d285b2ced98e2683a05a51fad73a627f05b6e81d68345d83ac.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "6e1075c04f2d38633d0064852191704c4b999b3977cd13af658ddee0bedba15d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "53830bdd3bfe8496330fedf9a2a49397593f50882a7c553e93aeb2be369d0a42", "generation_seconds_batch": 24.13657992298249, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0307748317718506, "target_z": 0.1972656581063975, "ce": 6.133810997009277, "genuineness": 4.337893962860107, "identity": 0.5940028429031372, "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_DRAM_reg/6e1075c04f2d38633d0064852191704c4b999b3977cd13af658ddee0bedba15d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "6e305074ab2c4a18db8720ede4c7df3bfe8cfa9d132666095d558b0bf4b9e869", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "bb35f65f442e2edc889617e6d8e5cdc5e661773cbb4112dfed4c22007f3f2370", "generation_seconds_batch": 15.30126704997383, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1331350803375244, "target_z": 0.2871094563354704, "ce": 6.308549880981445, "genuineness": 4.607763767242432, "identity": 0.6043267846107483, "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_DRAM_reg/6e305074ab2c4a18db8720ede4c7df3bfe8cfa9d132666095d558b0bf4b9e869.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "6e9c66b957e45ff952d2369ff118d719a2fba041179d5e653e7215bf2899e609", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "2ba46e226c1d0fd57e160e176425c09d8e8c9f54d19f1dbfb23f82ec8c3ce7d9", "generation_seconds_batch": 15.197967244079337, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8110342025756836, "target_z": 0.004394567728345402, "ce": 6.316492557525635, "genuineness": 4.6203155517578125, "identity": 0.6375530362129211, "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_DRAM_reg/6e9c66b957e45ff952d2369ff118d719a2fba041179d5e653e7215bf2899e609.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "6f03294566688d6bd72d97b37be73b17a34579b2690df5e493e59614165ebb41", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "aedfec117dacbf968fb60fff653407ed507d70564c4f3dfd09824d21c450d204", "generation_seconds_batch": 18.86060091597028, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.776224136352539, "target_z": 0.8515625931883781, "ce": 5.470598220825195, "genuineness": 4.532451152801514, "identity": 0.7961485385894775, "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_DRAM_reg/6f03294566688d6bd72d97b37be73b17a34579b2690df5e493e59614165ebb41.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "6f391f6d17e6256b9c53954829a1849b14be6603dd005001027a1bb4054ee6c6", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "fade48418ac006cebdb5034b93039f734eae51a3a8338f810ea48e81ebe5b8ff", "generation_seconds_batch": 21.312391487997957, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.918638229370117, "target_z": 0.9765625147139545, "ce": 5.617537498474121, "genuineness": 4.243753433227539, "identity": 0.26504477858543396, "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_DRAM_reg/6f391f6d17e6256b9c53954829a1849b14be6603dd005001027a1bb4054ee6c6.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": "6f9b33ff75c6380877cf6de1fbb2b37f32f2bbe90ec7dc195545865d9fe16e23", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Your message arrived after I had left.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9cbeff52942b672a8dcb6d156bc5646b9aeeb1c3b1e5ee613d760478d880c787", "generation_seconds_batch": 22.30722945393063, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.621808052062988, "target_z": 1.5937498365116172, "ce": 5.522225379943848, "genuineness": 3.798154592514038, "identity": 0.1946088671684265, "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_DRAM_reg/6f9b33ff75c6380877cf6de1fbb2b37f32f2bbe90ec7dc195545865d9fe16e23.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "6fee74616c86074a03ef973dae197b79608c9e33684b54842488abd3001df606", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "21bb379ef1b6a4c5b994da87b0b4411bdafb4f843421627d20013b7e5e154336", "generation_seconds_batch": 20.80017586098984, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8340797424316406, "target_z": 0.9023437066755786, "ce": 5.454712867736816, "genuineness": 4.444586753845215, "identity": 0.54986572265625, "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_DRAM_reg/6fee74616c86074a03ef973dae197b79608c9e33684b54842488abd3001df606.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "705af3bcac3cd315b377ec78d75d673c349f605283a3b2a1e1f12c60e1471c58", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "c7f111aab8e593eb2579b443ffa8d7d55f0d02ca71f941e9484595d35909231d", "generation_seconds_batch": 19.797749129938893, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.016548156738281, "target_z": 1.062500170027918, "ce": 5.692992687225342, "genuineness": 4.20609712600708, "identity": 0.514482319355011, "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": 0.03999999999999915, "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_DRAM_reg/705af3bcac3cd315b377ec78d75d673c349f605283a3b2a1e1f12c60e1471c58.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "70636774b3bbfb315273b637afe3cd923bc4887a9b79f57bb651900188cb54e6", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "fe245942e04d2e030a25d03f718c54d433e3cf260c0e7ddd55087b86e1f4515b", "generation_seconds_batch": 21.312391487997957, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.08775520324707, "target_z": 1.1250001307907063, "ce": 5.911416053771973, "genuineness": 4.281409740447998, "identity": 0.3992766737937927, "hypothesis": "encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "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_DRAM_reg/70636774b3bbfb315273b637afe3cd923bc4887a9b79f57bb651900188cb54e6.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "710d96263ecd6b769a87cfc977c463d386f0553d1429ca4df26fe8fec8c79488", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "f583b0340239aedc98eb1ffb2192cd53cef6ad492d8793622a95460c6dc1b683", "generation_seconds_batch": 21.121540046064183, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.998746156692505, "target_z": 1.046874970572091, "ce": 5.546053409576416, "genuineness": 4.168440818786621, "identity": 0.4494284987449646, "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_DRAM_reg/710d96263ecd6b769a87cfc977c463d386f0553d1429ca4df26fe8fec8c79488.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "7217cc74980effcf944c1fec3104e56d74584928c8b658aacbb0f68ae5f61a69", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "f60e575ca700847b22ea7110c68c2d3666a2aa3cc1a165bd014874db6eba49e7", "generation_seconds_batch": 20.330678243073635, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.061052322387695, "target_z": 1.1015624362395307, "ce": 6.308549880981445, "genuineness": 4.268857479095459, "identity": 0.12973274290561676, "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_DRAM_reg/7217cc74980effcf944c1fec3104e56d74584928c8b658aacbb0f68ae5f61a69.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "72496569a699904bdda0e58fd66fb22ae1fd4c75abdc600754ef39a84edb690c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "e39aa23784a4a4b32c2ebc350e340dd129a360f6dbb8ca82aeb1615c3bfae018", "generation_seconds_batch": 20.001645853044465, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4913957118988037, "target_z": 0.6015625408720957, "ce": 6.054384231567383, "genuineness": 4.545003414154053, "identity": 0.44946199655532837, "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_DRAM_reg/72496569a699904bdda0e58fd66fb22ae1fd4c75abdc600754ef39a84edb690c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "72a5aeb4e91a11d82a41d09db0e5a99815ca96535cc2f53aec43b02ace76ab1d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "82b9228e01dc8c353b12e77f29872f02278d3984228e4fc13d4bd626bd340b3a", "generation_seconds_batch": 22.0043279719539, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5002965927124023, "target_z": 0.6093750359674442, "ce": 5.998785495758057, "genuineness": 4.545003414154053, "identity": 0.5883016586303711, "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_DRAM_reg/72a5aeb4e91a11d82a41d09db0e5a99815ca96535cc2f53aec43b02ace76ab1d.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": "72a80739535c671860ed12c0fb40cc462ff36381423c3996e877932aa0dca864", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The road turns left behind the bridge.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "ee3fe3025348c178d78e6ad6dc1b2b8f83161a14921664e40c2579f2b155285d", "generation_seconds_batch": 19.970064062974416, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.497195720672607, "target_z": 1.484374905176738, "ce": 5.788304805755615, "genuineness": 3.785602569580078, "identity": 0.34208470582962036, "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_DRAM_reg/72a80739535c671860ed12c0fb40cc462ff36381423c3996e877932aa0dca864.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "72aec2a962753a3460cef8bb1a2f651be660a983eadd0d87a97371f701e80c7b", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "c5490fc4267278457a19675fca78eb85efb590dc79faa9796ce6eeea83f6c53c", "generation_seconds_batch": 20.70685949001927, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.8621320724487305, "target_z": 1.8046874133511572, "ce": 5.363371849060059, "genuineness": 3.8797430992126465, "identity": 0.10249175876379013, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude graam. We kunnen er samen in kijken.", "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_DRAM_reg/72aec2a962753a3460cef8bb1a2f651be660a983eadd0d87a97371f701e80c7b.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "730f0c75d8fc6ff5de160206cb61359d320ecc5bb167fce21008b488d7a6b114", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "6eefb082447eafc093ca3990a13bddb0839ec09e12236e2a32b272570d24ea4e", "generation_seconds_batch": 15.446222542086616, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7139177322387695, "target_z": 0.7968749182558086, "ce": 5.1350202560424805, "genuineness": 4.331617832183838, "identity": 0.3051224946975708, "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_DRAM_reg/730f0c75d8fc6ff5de160206cb61359d320ecc5bb167fce21008b488d7a6b114.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": "748d0ce4292cb7203f488f86728a8e5e356afbd81fb2b33e6f08ef1fa91acc41", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "603064ae16a5d84fb852f5efbd0c4ec8e2f9518ccf1154bbb022330dc02a85df", "generation_seconds_batch": 22.50282762304414, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.47939395904541, "target_z": 1.468749914986041, "ce": 5.526196479797363, "genuineness": 4.005263805389404, "identity": 0.6913968324661255, "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_DRAM_reg/748d0ce4292cb7203f488f86728a8e5e356afbd81fb2b33e6f08ef1fa91acc41.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "74c956507c2d89dd1454d481e7dae3a56f8a8788585c54bd313a4636319be256", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "559533bea20e5807ddc1a9d3f97b1984cab1eea52b53f62de2e0714b1c993b74", "generation_seconds_batch": 21.377519707079045, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.705016851425171, "target_z": 0.7890624231604602, "ce": 5.546053409576416, "genuineness": 4.2374773025512695, "identity": 0.3951089382171631, "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_DRAM_reg/74c956507c2d89dd1454d481e7dae3a56f8a8788585c54bd313a4636319be256.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "74ea45844c9410a2d0723fa3e205496b947ba8861283ab69e411cde770a2bade", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "f967247e9c94cba3107021ace6a8205861f94726118802b36ea449efc3166de8", "generation_seconds_batch": 21.68604837998282, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6160080432891846, "target_z": 0.7109374722069749, "ce": 5.244231700897217, "genuineness": 4.3127899169921875, "identity": 0.3255683183670044, "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_DRAM_reg/74ea45844c9410a2d0723fa3e205496b947ba8861283ab69e411cde770a2bade.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "7592879d7f839499f8713828ae746231051ee393e208bac7368e4d4dfba79ede", "bucket": "vn_S_DRAM_reg", "phase": "wording", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "1b52201692bfbdc1c8491f62ed1a27680805a994993470ce6fb619b0e34528ae", "generation_seconds_batch": 14.078537620953284, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1843152046203613, "target_z": 0.3320313554500069, "ce": 6.197352409362793, "genuineness": 4.080576419830322, "identity": 0.32928919792175293, "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_DRAM_reg/7592879d7f839499f8713828ae746231051ee393e208bac7368e4d4dfba79ede.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "75afc3f6027d7e79a6660eb09421a5d4396d104e4889db8aebf8615658df4a19", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "e0b1c967a8c28af0c87a73a0428b0b5a8374022c149d60fcf212255b35e612ac", "generation_seconds_batch": 21.34764486597851, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.461592197418213, "target_z": 1.453124924795344, "ce": 5.470598220825195, "genuineness": 4.005263805389404, "identity": 0.30691319704055786, "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_DRAM_reg/75afc3f6027d7e79a6660eb09421a5d4396d104e4889db8aebf8615658df4a19.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "764bdb07c056fc90d6c404f43e7b37875e889f96589c29db0f1324517234e5ab", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "dbce227918356c253a66ef6ede20726180e23c7a3296ac4d3848b62bc6d74b91", "generation_seconds_batch": 20.61641469306778, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4277396202087402, "target_z": -0.3320313554500069, "ce": 5.839932441711426, "genuineness": 4.557555198669434, "identity": 0.7471911907196045, "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_DRAM_reg/764bdb07c056fc90d6c404f43e7b37875e889f96589c29db0f1324517234e5ab.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "764c86b17794d591a5a329f1cc5933a80580e4cf7ce8a7282b858274537ebe21", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "5e8faae5e2a7be808fc8ee7ef3cc649f87400360254bc7ca911b21651f7bf80b", "generation_seconds_batch": 20.619994111009873, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.900599241256714, "target_z": 0.08300773422993676, "ce": 6.109982967376709, "genuineness": 3.998987913131714, "identity": 0.43735647201538086, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/764c86b17794d591a5a329f1cc5933a80580e4cf7ce8a7282b858274537ebe21.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "76cb245d7cfc275d5ed2f4cc638ad10324437486e6e5a6fda34f672a2253e842", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "48805f547aa82bd55d069e210fcf8e4a2e902f99ab3625d31df43eb5fc19964e", "generation_seconds_batch": 21.39462256198749, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.750953197479248, "target_z": -0.048339826481539575, "ce": 6.1894097328186035, "genuineness": 4.080576419830322, "identity": 0.7560964822769165, "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_DRAM_reg/76cb245d7cfc275d5ed2f4cc638ad10324437486e6e5a6fda34f672a2253e842.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "76e439d15beac86f644fb1723458f9f09295962b751ff1605588aa1b839ffa3e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "a557c0fb669078f66b9616408d1f7546d7e2de4b18717f40129653c2e161eba0", "generation_seconds_batch": 15.955625968053937, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6413609981536865, "target_z": -0.14453126389651252, "ce": 6.268836498260498, "genuineness": 4.444586753845215, "identity": 0.5284817814826965, "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_DRAM_reg/76e439d15beac86f644fb1723458f9f09295962b751ff1605588aa1b839ffa3e.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "774255946b89ad9649c873261ef2f62f641cf37ec12e562daf2aa8876a8a8e0e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "9950d22e224aa8143f33204ac02156381ac1d8bba99c3e5a28eec4bafabf5553", "generation_seconds_batch": 15.729167327983305, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.8888349533081055, "target_z": 1.8281251079023326, "ce": 6.141753673553467, "genuineness": 3.911123275756836, "identity": 0.4091988801956177, "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_DRAM_reg/774255946b89ad9649c873261ef2f62f641cf37ec12e562daf2aa8876a8a8e0e.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "fb00f6ffb171939fb0a835a211505a81421d7080a8e8fb30009fd2e562a6a98b", "generation_seconds_batch": 16.15020708704833, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0263242721557617, "target_z": 0.1933593059261583, "ce": 6.221180438995361, "genuineness": 4.494794845581055, "identity": 0.8078892230987549, "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_DRAM_reg/77962898fb95ddc0e92a153772191fe01482e40f98d2245158cb92652e8a93c1.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "7808e31e8b7eda0c7f75423cc36878eda5cf6c5ee87dc9e2d1aa6652861f9a3b", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "140082e9c122b27ae4b2ff4668e6145b54df9aa0e83d023863de9b387b948afa", "generation_seconds_batch": 21.29505303595215, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7676422595977783, "target_z": -0.033691476652184825, "ce": 6.403861999511719, "genuineness": 4.582659721374512, "identity": 0.5453096628189087, "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_DRAM_reg/7808e31e8b7eda0c7f75423cc36878eda5cf6c5ee87dc9e2d1aa6652861f9a3b.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "780d128985d15b3d4f000ef147713af974411ecaa22f8acde9688ae1b769fd9b", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "61ca3e4271d62ff0a00337d60b84bcc0109e361a9d02fa41ec4b046c7aae58da", "generation_seconds_batch": 14.675983360037208, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2096681594848633, "target_z": -0.5234373806534806, "ce": 6.300607204437256, "genuineness": 4.6956281661987305, "identity": 0.4475579261779785, "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_DRAM_reg/780d128985d15b3d4f000ef147713af974411ecaa22f8acde9688ae1b769fd9b.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "782c507ccc189fa667a64ae3390645cd1ccace85f50c20c5dfdb2a979104e6ad", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "646afd02d8c325de10770647669684881cf0f21a6982ca12f393b66cc69b55c0", "generation_seconds_batch": 20.985833944985643, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.03434944152832, "target_z": 1.0781247416883553, "ce": 6.09409761428833, "genuineness": 4.419482707977295, "identity": 0.5694443583488464, "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": 0.03999999999999915, "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_DRAM_reg/782c507ccc189fa667a64ae3390645cd1ccace85f50c20c5dfdb2a979104e6ad.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": "78cd519698d8bfed0d32a89f53fa208f4844d4b213d2d712a31183eea3624f2b", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c04fffd32428bc8586409d6992477f0b8f03c1d3b3aadf90a543d13447354d58", "generation_seconds_batch": 22.243346164003015, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.874133825302124, "target_z": 0.9375000392372118, "ce": 5.220403671264648, "genuineness": 4.281409740447998, "identity": 0.7743529081344604, "hypothesis": "Les gliss sont sur la table de la cuisine. J'ai regardé dans le tiroir, puis près de la fenêtre. Prends-les quand tu partiras.", "wer": 0.041666666666666664, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_DRAM_reg/78cd519698d8bfed0d32a89f53fa208f4844d4b213d2d712a31183eea3624f2b.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "78f38aa0c57accf367b2f3b0a86b79d33b2e840610a53fe0bc62fa1ea1246a5a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "5f10cebb03d60d7e855425e3426953fffab8cd2c67e2f45a92c19fe86b9df151", "generation_seconds_batch": 21.857483603991568, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.977843761444092, "target_z": 1.9062500588558178, "ce": 5.601652145385742, "genuineness": 3.823258876800537, "identity": 0.4383871555328369, "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_DRAM_reg/78f38aa0c57accf367b2f3b0a86b79d33b2e840610a53fe0bc62fa1ea1246a5a.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "7924af47e7f093942cd3a92d1757787a46fc580e9a716c9ba7c7fc5c9e42775a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "b2a555f083b257d23a887443a09a7248f3f46825f9e6fabe76f2d2ba2724f8ea", "generation_seconds_batch": 18.78322666697204, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.758422374725342, "target_z": 0.8359376029976812, "ce": 4.958295822143555, "genuineness": 4.18726921081543, "identity": 0.025748446583747864, "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": 0.03999999999999915, "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_DRAM_reg/7924af47e7f093942cd3a92d1757787a46fc580e9a716c9ba7c7fc5c9e42775a.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "ae99717eb15ae2791d3aaa72849a4de4e0dc4e8cea82f23ffc8688089c3c5d91", "generation_seconds_batch": 15.823777557932772, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2831003665924072, "target_z": -0.4589843484331378, "ce": 6.086154937744141, "genuineness": 4.406930446624756, "identity": 0.6209487318992615, "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_DRAM_reg/794d3b9296e3c4be7c1e4182a2fa3cf095f7d593990763ce0f19c1f5e499a989.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": "795ba003c5efd827ff417bbc531fee50d1fd99abbcd354f44d7de7c7454be9c9", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3cee90653133f0f9a7539f7656a6884230c898b7324e56539b7de132c6bb4e87", "generation_seconds_batch": 22.093182004988194, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.998746156692505, "target_z": 1.046874970572091, "ce": 5.3514580726623535, "genuineness": 4.268857479095459, "identity": 0.6306043267250061, "hypothesis": "Quelqu'un a déplacé la petite chasse bleue ! Elle était ici, près de la grande bibliothèque. Nous la remettrons après le déjeuner.", "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_DRAM_reg/795ba003c5efd827ff417bbc531fee50d1fd99abbcd354f44d7de7c7454be9c9.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "79abdd4a2bbd5e47fc553e55e6b3e1889b895ad71c33118050c6ee82fe8e8358", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "e96af4cab7feace06db5feb2e2750525fa668eec449d30772bdae4e1d44cc182", "generation_seconds_batch": 21.20461635896936, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.25997257232666, "target_z": 0.3984374591279043, "ce": 5.736677646636963, "genuineness": 4.344170093536377, "identity": 0.39732155203819275, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/79abdd4a2bbd5e47fc553e55e6b3e1889b895ad71c33118050c6ee82fe8e8358.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "7a16840e0e2f26a709150b4e7cb8d1d3023d82c6659e16b8d88810f361e749f1", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "2dda749ed0b5b5954e0b69226ab335febb82b056f214eb77b039e190dbbf2bde", "generation_seconds_batch": 21.13125059194863, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7228188514709473, "target_z": 0.804687622616287, "ce": 6.157639026641846, "genuineness": 4.325342178344727, "identity": 0.7411742806434631, "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": 0.03999999999999915, "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_DRAM_reg/7a16840e0e2f26a709150b4e7cb8d1d3023d82c6659e16b8d88810f361e749f1.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "7a4789de2b3c90647a4efe39eb941dd4ba7a0a2a34f0d25e2e278c78755c4ada", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "2230faf4ebd11a7fa2a614f3a369a7284674e2689fb4b4c36b71b179f8f635e6", "generation_seconds_batch": 20.972861030953936, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7940258979797363, "target_z": 0.8671875833790752, "ce": 6.23706579208374, "genuineness": 4.218649387359619, "identity": 0.7359374761581421, "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_DRAM_reg/7a4789de2b3c90647a4efe39eb941dd4ba7a0a2a34f0d25e2e278c78755c4ada.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "7a4bc4373041f71bdaa11bd4cbf69093354f9d9ba03ad497b4db9dc64ef4e6aa", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "940eec6aa8fb885aa3a1e2b27ee3803704c078d7b1b74ad5e82f2de2f3bfc35d", "generation_seconds_batch": 21.190073181991465, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 5.565301895141602, "target_z": 2.42187473514882, "ce": 5.696964263916016, "genuineness": 3.6914620399475098, "identity": 0.5624403953552246, "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_DRAM_reg/7a4bc4373041f71bdaa11bd4cbf69093354f9d9ba03ad497b4db9dc64ef4e6aa.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": "7a5333fb9e573713c1a00cbcff1a57a3f1675b7864e9125f57fcc7adee8417d8", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "93e9683f60dab86737460ecc7e4ab645cc0897987fcb248565cb11720c1c419c", "generation_seconds_batch": 15.471985918004066, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7406206130981445, "target_z": 0.8203126128069841, "ce": 5.967014789581299, "genuineness": 4.268857479095459, "identity": 0.42188963294029236, "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_DRAM_reg/7a5333fb9e573713c1a00cbcff1a57a3f1675b7864e9125f57fcc7adee8417d8.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "7ad4e45ce69bcd2486f673c6451ba67a1af1eb6b9cd71fe3d635dab6de4dc779", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "f4375e4c99a04547cb7b64a7f76eeecbecd1e6edc26cea28766ad699029ce9d1", "generation_seconds_batch": 16.149571178946644, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2555222511291504, "target_z": 0.394531316212795, "ce": 6.2767791748046875, "genuineness": 4.582659721374512, "identity": 0.5922609567642212, "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_DRAM_reg/7ad4e45ce69bcd2486f673c6451ba67a1af1eb6b9cd71fe3d635dab6de4dc779.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "7adf1d6a6b1f83fcb1381ea7dc6bc5bc5ae4febb301747ac3b7a0d21cbdc2e60", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "5d8545d0f8aa333d66c994848e03af023eb4ea519a32e6363481952b94c0f782", "generation_seconds_batch": 16.104008362977766, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.194565773010254, "target_z": 1.2187500719348885, "ce": 5.629451274871826, "genuineness": 3.986435890197754, "identity": 0.503569483757019, "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_DRAM_reg/7adf1d6a6b1f83fcb1381ea7dc6bc5bc5ae4febb301747ac3b7a0d21cbdc2e60.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": "7b8400ef046652acee93618a534edb824a31786722bd26875eaf1d133d3b97ec", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "e1ee328e20a4ded7c52537ec3fb31bfcae2072e1d09be06ca8c7c46ce58304b3", "generation_seconds_batch": 15.53555332601536, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9097373485565186, "target_z": 0.968750019618606, "ce": 5.9431867599487305, "genuineness": 3.9299514293670654, "identity": 0.5981476306915283, "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_DRAM_reg/7b8400ef046652acee93618a534edb824a31786722bd26875eaf1d133d3b97ec.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "7b9f076ec917cc190161ea2ecb8efea37a70018f38de32fa809022fbf9c8ba40", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "97136def91d9aeb662b92ab9ded31d5db43a772296aef7b948db4c9ac395818c", "generation_seconds_batch": 20.26349670998752, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0841801166534424, "target_z": 0.24414062867848862, "ce": 5.911416053771973, "genuineness": 4.137060642242432, "identity": 0.5634193420410156, "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_DRAM_reg/7b9f076ec917cc190161ea2ecb8efea37a70018f38de32fa809022fbf9c8ba40.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "7baaa3052ff9cb4bc6f7fbffbdebc28dbbf73eec5498b89035f2153b0601e9ef", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "be9f6ae23dfa1dafcac4ecdf87dc26813241dece6366dd3698bb3e98777d82eb", "generation_seconds_batch": 21.549519784050062, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1509368419647217, "target_z": 0.30273444652616743, "ce": 6.205295085906982, "genuineness": 4.494794845581055, "identity": 0.706413984298706, "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": 0.03999999999999915, "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_DRAM_reg/7baaa3052ff9cb4bc6f7fbffbdebc28dbbf73eec5498b89035f2153b0601e9ef.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "7bdaa98be24fd5f2249ed7a5381697b42574fc371489a42ce3aa5ee0597a9f17", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "14649504ccdad83bd0bdb55bff19ab9cb9b98e412f756ff8a671a8f616b3a926", "generation_seconds_batch": 20.576529294019565, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.727269172668457, "target_z": 0.8085937655313964, "ce": 5.502368927001953, "genuineness": 4.419482707977295, "identity": 0.5734590888023376, "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_DRAM_reg/7bdaa98be24fd5f2249ed7a5381697b42574fc371489a42ce3aa5ee0597a9f17.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "7bea79a401de1612de27c7494836556612f6b6bb32fb8fe1d989700902c69271", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "d2887fd4a6ceda2b43618d1307dd0888a1230c01814c30f38390967bf9bedf14", "generation_seconds_batch": 20.274366572964936, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8894731998443604, "target_z": 0.07324216767703359, "ce": 6.205295085906982, "genuineness": 4.68307638168335, "identity": 0.5551801323890686, "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_DRAM_reg/7bea79a401de1612de27c7494836556612f6b6bb32fb8fe1d989700902c69271.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "7c0d29d77eb81eb80b9ae504702573a04085f1c3d06c9352af502ef080eecdff", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "c5278babe91b8ef5f7865826c9ad99e747b66aa033ddd3b4e3ec1d2ad3338248", "generation_seconds_batch": 20.375034737982787, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.333405017852783, "target_z": 0.462890700613377, "ce": 6.0305562019348145, "genuineness": 4.494794845581055, "identity": 0.486918568611145, "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_DRAM_reg/7c0d29d77eb81eb80b9ae504702573a04085f1c3d06c9352af502ef080eecdff.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "7c1f73b40d00fe989334356f488be7f1c2da8e037735e6ac52237d85607b893b", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "80908efccd4d8465417baae68655b684af350bf22fa8487230a2d56fb35d9161", "generation_seconds_batch": 21.03179468109738, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7626357078552246, "target_z": -0.0380858351154003, "ce": 6.229123115539551, "genuineness": 4.519899368286133, "identity": 0.6316273212432861, "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_DRAM_reg/7c1f73b40d00fe989334356f488be7f1c2da8e037735e6ac52237d85607b893b.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "7c273da81d4e0813a4d6d85c213913dfce4b5ebcb024a814841f7178cd6c5816", "bucket": "vn_S_DRAM_reg", "phase": "wording", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "51b8258da10af1dd4861ca863fe51aed2eb21be3f7249b94396230585f78f628", "generation_seconds_batch": 15.074661899008788, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4967215061187744, "target_z": -0.2714844661447734, "ce": 6.252951145172119, "genuineness": 4.607763767242432, "identity": 0.48755207657814026, "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_DRAM_reg/7c273da81d4e0813a4d6d85c213913dfce4b5ebcb024a814841f7178cd6c5816.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": "7c7cfe4c09df1fe7c6ee8a8add56d4a6c693ece2cec6c494754e0b619c70b17c", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1a4b41f2f9d43fa6dfa34e4e4f84ba39b4973d9c80f2f7d7a0662dd3a7fb71d3", "generation_seconds_batch": 22.232289623003453, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.399286270141602, "target_z": 1.3984376683930342, "ce": 5.391171455383301, "genuineness": 3.7918784618377686, "identity": 0.030838068574666977, "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_DRAM_reg/7c7cfe4c09df1fe7c6ee8a8add56d4a6c693ece2cec6c494754e0b619c70b17c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "7c8b70db29ad780e353c3be2633542efd527d6aa34c4c5650f20ba208ae10b07", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "eda4917e82874b66d53029e22c18020ffd6942d36882a79a7b04743e59ba5694", "generation_seconds_batch": 21.647308001993224, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.610208034515381, "target_z": -0.17187489209766738, "ce": 6.09409761428833, "genuineness": 4.419482707977295, "identity": 0.8017027378082275, "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_DRAM_reg/7c8b70db29ad780e353c3be2633542efd527d6aa34c4c5650f20ba208ae10b07.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "7cde80e05b5e90eaed86d3996153372a9f948da5bafee33dc73d932175822ce8", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "7d7001b321da7c8f77447850fd49102496a8190d3b25d7d79f6780f1434b375a", "generation_seconds_batch": 16.297408030950464, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2710986137390137, "target_z": 0.40820302568080746, "ce": 6.181467056274414, "genuineness": 4.413206577301025, "identity": 0.6002633571624756, "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_DRAM_reg/7cde80e05b5e90eaed86d3996153372a9f948da5bafee33dc73d932175822ce8.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "7cef2d97a8921f13c1216e0ec5ac5b67c14d532bfe0080a2b11543dcb177a996", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "2ebdb6287151e0815e0c76db8ce3479ab6b1e6902984648735c523dbbbac1ed8", "generation_seconds_batch": 19.633071449003182, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.604006290435791, "target_z": 1.5781248463209203, "ce": 5.744620323181152, "genuineness": 3.9676074981689453, "identity": 0.32488536834716797, "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_DRAM_reg/7cef2d97a8921f13c1216e0ec5ac5b67c14d532bfe0080a2b11543dcb177a996.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "7d0fec99cd275b883a6eaa2a4584bfcc618883f0e481fa1627f286e0a91fae0f", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "7f82ac7076a10cdf62179dfdc7dd207acaa8fe737c8a5e401bae0009d156cff5", "generation_seconds_batch": 20.55909958994016, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.406837224960327, "target_z": 0.5273437328337198, "ce": 6.403861999511719, "genuineness": 4.3504462242126465, "identity": 0.3888697028160095, "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_DRAM_reg/7d0fec99cd275b883a6eaa2a4584bfcc618883f0e481fa1627f286e0a91fae0f.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": "7e68839d823aafcc118a36e7bc05ef44735eacb11a4c3c76a49d05602b72a021", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "2fbc80d3239650cac4759c43d3bd066c26960f4ef05d1236c419dc067e87817f", "generation_seconds_batch": 15.53555332601536, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6691763401031494, "target_z": -0.12011713824912469, "ce": 6.173524379730225, "genuineness": 4.394378662109375, "identity": 0.7098516225814819, "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_DRAM_reg/7e68839d823aafcc118a36e7bc05ef44735eacb11a4c3c76a49d05602b72a021.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "7f0e9653129535ef0c5ee9e9c190bd71d49857dd6e518619ec75753bd1b71bfd", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "66293751a57fc359625b81861620b6c09af4713322b141c11b071526b334d03a", "generation_seconds_batch": 14.900078602018766, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5448009967803955, "target_z": 0.6484375114441868, "ce": 5.792276382446289, "genuineness": 4.595211505889893, "identity": 0.7666206955909729, "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_DRAM_reg/7f0e9653129535ef0c5ee9e9c190bd71d49857dd6e518619ec75753bd1b71bfd.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "c7961b28acb184e36d11552f9422beb0141b7eb99a472b8beed071d791e743a5", "generation_seconds_batch": 15.31716720201075, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8536325693130493, "target_z": -0.8359374983651162, "ce": 6.395919322967529, "genuineness": 4.406930446624756, "identity": 0.2673768997192383, "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_DRAM_reg/7f6fce1a6f733719f45a363f28fee4e24b2f07a3396aa07b42392913d7c24239.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "7f8ecff8415889f94c8d0c7c2845dafc8d6e70ec67e851f3fac9fe8da9d6835d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "fb0b8ce9e8070d447e31056c3c1d103cb6b82d1161306b082fe2ee311e710075", "generation_seconds_batch": 21.5447271540761, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.657411575317383, "target_z": 1.6249998168930113, "ce": 5.470598220825195, "genuineness": 3.9550557136535645, "identity": 0.4474327266216278, "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_DRAM_reg/7f8ecff8415889f94c8d0c7c2845dafc8d6e70ec67e851f3fac9fe8da9d6835d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "7fdbcaf4707b45115d68349f7cca93d739abb3042a08ba29a3e2ad9c90fef2b4", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "50bf91dc28a8d2a399ada871b80a93d071eaaa4b9496a492bce1d0cf160d8f97", "generation_seconds_batch": 20.547909564105794, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.443790435791016, "target_z": 1.4374999346046469, "ce": 5.764476776123047, "genuineness": 4.231201171875, "identity": 0.525083601474762, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/7fdbcaf4707b45115d68349f7cca93d739abb3042a08ba29a3e2ad9c90fef2b4.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "80153d3804a4fcbc5362eee978dff73f0d8c98f53d749af15c971c7052486381", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "5a32162b0736c46f62f80803e083864337b42cbd3053694746e0523a5f0385d2", "generation_seconds_batch": 16.424036265001632, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.553701877593994, "target_z": 0.6562500065395354, "ce": 6.09409761428833, "genuineness": 4.419482707977295, "identity": 0.5000191926956177, "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_DRAM_reg/80153d3804a4fcbc5362eee978dff73f0d8c98f53d749af15c971c7052486381.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "8029c9bf2372ab75ae96390a0309ee543298d08d2080df77fc7f324c16ad7f07", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "8fcf34fbe70e01bb960e446ebc00b8b6cf8b9cf1eb2bc336a6da1acfab3a753d", "generation_seconds_batch": 21.856821782072075, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4722442626953125, "target_z": -0.29296867070813437, "ce": 6.070269584655762, "genuineness": 4.413206577301025, "identity": 0.630158543586731, "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_DRAM_reg/8029c9bf2372ab75ae96390a0309ee543298d08d2080df77fc7f324c16ad7f07.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "8086bb8c028c2ccbf38063d1307a90bbecb335dd892cf85d55747f3701259c3b", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "fb6ade8aeedb9a3def6ba0ec32cf76f5360db3187e3540b653701beecd9e20aa", "generation_seconds_batch": 15.147895290981978, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9631426334381104, "target_z": 1.015624990190697, "ce": 6.0146708488464355, "genuineness": 4.507347106933594, "identity": 0.8128559589385986, "hypothesis": "J'ai trouvé la petite boîte. Elle était près, elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.15789473684210525, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_DRAM_reg/8086bb8c028c2ccbf38063d1307a90bbecb335dd892cf85d55747f3701259c3b.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "80b213c7d8bae2b24e8f93db41c31d3155399514455e2c1ddeb4bd7f90151c4c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "abdbb378233870d2c7c010d819a34ed9ee82dff49bede2ce08db7a6e54b2f646", "generation_seconds_batch": 21.150260988040827, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2541723251342773, "target_z": -0.4843751144418679, "ce": 5.959072113037109, "genuineness": 4.15588903427124, "identity": 0.8168042898178101, "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_DRAM_reg/80b213c7d8bae2b24e8f93db41c31d3155399514455e2c1ddeb4bd7f90151c4c.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "88151b81e7b1d30cffa1111df49090aabbc094e6d2d52f949815be8596375dd1", "generation_seconds_batch": 16.395268750027753, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3543074131011963, "target_z": -0.39648438767034966, "ce": 6.372091293334961, "genuineness": 4.582659721374512, "identity": 0.5862210988998413, "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_DRAM_reg/80c2597a6b8f43b5338a6b37d654868d679f46ded48514b42837ab18fba327ad.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "80dd46b62c76167e113a1222f6d9d69a955628d475e0955da120f99f3ed29e85", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "2a628c4daa2cce32c04699e9413d21e3780495346df192985c0e019cd9f30029", "generation_seconds_batch": 22.183748372015543, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3756842613220215, "target_z": 0.500000104632565, "ce": 6.284721851348877, "genuineness": 3.886019229888916, "identity": 0.301474928855896, "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_DRAM_reg/80dd46b62c76167e113a1222f6d9d69a955628d475e0955da120f99f3ed29e85.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "810099c3e08658ec5b1d6e181d0317a8757b946d14e32cf20735b640d30511d2", "bucket": "vn_S_DRAM_reg", "phase": "wording", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "167e341de9eda0bac7a1f2093a217525115d49a1235fdf74c76f6f9a966c93ee", "generation_seconds_batch": 15.908091485034674, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6124331951141357, "target_z": -0.16992182064011274, "ce": 6.1894097328186035, "genuineness": 4.494794845581055, "identity": 0.7317900657653809, "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_DRAM_reg/810099c3e08658ec5b1d6e181d0317a8757b946d14e32cf20735b640d30511d2.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "815770b819c3015663b4bd2fcff253326728d13bb819eca093575448f517aed8", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "06a119113c565af8ba455f28297779fd951b607b639daa34f1891b6d843a0e82", "generation_seconds_batch": 20.564424208016135, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.847431182861328, "target_z": 0.9140625539511663, "ce": 6.157639026641846, "genuineness": 4.218649387359619, "identity": 0.3963088393211365, "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_DRAM_reg/815770b819c3015663b4bd2fcff253326728d13bb819eca093575448f517aed8.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "815e7419f62e02a4fb56a8913d78e107a1242ae50b481d215ca619b4f9b8f1d9", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "5ce1504260e2c467136f06518b5e6c5ce9185a5f6ce4ab65cea103bb10d8fd8d", "generation_seconds_batch": 21.121540046064183, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8652329444885254, "target_z": 0.9296875441418634, "ce": 5.335572719573975, "genuineness": 4.143336772918701, "identity": 0.18173342943191528, "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_DRAM_reg/815e7419f62e02a4fb56a8913d78e107a1242ae50b481d215ca619b4f9b8f1d9.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "81b5393ce85918829ae89c9cf9140144131e39e33d2e3a3131d88f6d0767c0f3", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "3fa0e418414004986ffd8b7bf9375896d55063664028c22ce07ba6aaebeb4247", "generation_seconds_batch": 21.198350839084014, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9631426334381104, "target_z": 1.015624990190697, "ce": 5.2640886306762695, "genuineness": 4.256305694580078, "identity": 0.4388006925582886, "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_DRAM_reg/81b5393ce85918829ae89c9cf9140144131e39e33d2e3a3131d88f6d0767c0f3.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "81d53b5ac03e9bc9174a603135914a88fa57ccd45dc65c40de1f47f828b02de3", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "f19c3dcbed61422142c5b0a330af9b5a36e46adac23e887ac9c211c7bb8d3020", "generation_seconds_batch": 21.2666702340357, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.874133825302124, "target_z": 0.9375000392372118, "ce": 6.109982967376709, "genuineness": 4.18099308013916, "identity": 0.7655484676361084, "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": 0.03999999999999915, "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_DRAM_reg/81d53b5ac03e9bc9174a603135914a88fa57ccd45dc65c40de1f47f828b02de3.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "8236874e114b7b9a90a3c5ad3825e712a39353b0de8cd03e41166e1708f48e29", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "088546da6d2d9dd13cce1c1ad664c0cb7a2237eff3e0cec2a9179736cb3ce16b", "generation_seconds_batch": 22.83756926190108, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.503397226333618, "target_z": -0.2656250425069795, "ce": 6.038498878479004, "genuineness": 4.645420074462891, "identity": 0.6688295006752014, "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_DRAM_reg/8236874e114b7b9a90a3c5ad3825e712a39353b0de8cd03e41166e1708f48e29.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "825230709d358c4ddff62c61433ab8ce2321b60dc495cc61cf043080ab11a8db", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "d585519220af49d9d31123d89f6a80a92cd9a3b70d4629e8ee81ddc7fea292e5", "generation_seconds_batch": 22.134722555987537, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3289544582366943, "target_z": 0.4589843484331378, "ce": 6.078212261199951, "genuineness": 4.344170093536377, "identity": 0.34736090898513794, "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": 0.03999999999999915, "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_DRAM_reg/825230709d358c4ddff62c61433ab8ce2321b60dc495cc61cf043080ab11a8db.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "82e0324dce4e093dc983f5a1394be1e9c54b8d8a3c127907df22da79bf2d8200", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "ec41c73e89f5ec2038af85e29450c40b9f8db24a5c1b707be7693d44cfe3fca9", "generation_seconds_batch": 21.597625555936247, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.406837224960327, "target_z": 0.5273437328337198, "ce": 5.772419452667236, "genuineness": 4.607763767242432, "identity": 0.71098792552948, "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_DRAM_reg/82e0324dce4e093dc983f5a1394be1e9c54b8d8a3c127907df22da79bf2d8200.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "83f8973b868a0e52d678b63632ee53cb2d9931a63815eb3d3efa586e09077e6a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "ed42a400af992cee363a4c312d0460f497d5765e7db3bcb734132e930a18fead", "generation_seconds_batch": 20.773486281977966, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.2746734619140625, "target_z": 1.289062318527895, "ce": 5.065521717071533, "genuineness": 4.168440818786621, "identity": 0.2858317196369171, "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_DRAM_reg/83f8973b868a0e52d678b63632ee53cb2d9931a63815eb3d3efa586e09077e6a.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "840e61d0e2de5489a61f1b2d1af33f48a7a5efe300ba6363bf5675f3c4315fa7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "96d47ce0ed5a6a64ea06de4516bfe01af9c49bf67a3beafc1eeea9017d81e527", "generation_seconds_batch": 20.62978708196897, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.758422374725342, "target_z": 0.8359376029976812, "ce": 6.1020402908325195, "genuineness": 4.482243061065674, "identity": 0.7207921743392944, "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_DRAM_reg/840e61d0e2de5489a61f1b2d1af33f48a7a5efe300ba6363bf5675f3c4315fa7.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "842c9fd8ed2c655f42c4193f595bec8911d1a07f29b6a7694f29fcf5c5bb8674", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "394edb92bc504f2a693e3a7330c7d9d0a4ffe91d43d1be736b1ead81224c12c7", "generation_seconds_batch": 15.397843602928333, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.776224136352539, "target_z": 0.8515625931883781, "ce": 6.356205940246582, "genuineness": 4.369274139404297, "identity": 0.4758327007293701, "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_DRAM_reg/842c9fd8ed2c655f42c4193f595bec8911d1a07f29b6a7694f29fcf5c5bb8674.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "844531be2120a9957d3c86072d5901212eb2e9f501d816f630803639fdb9a46e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "64fb7ff03a91abb1d36380c70baaa00977187664a9a15d2ea4d4bdb31b10439d", "generation_seconds_batch": 20.330678243073635, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.408186912536621, "target_z": 1.406249954223253, "ce": 5.824047088623047, "genuineness": 4.256305694580078, "identity": 0.28098028898239136, "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_DRAM_reg/844531be2120a9957d3c86072d5901212eb2e9f501d816f630803639fdb9a46e.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "84b47dc7488cdf559b5750abdd4ee6971081af343ae3af3ff3dc7d6c1f7955be", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "aba0ee960d4990e3efe3d42b473a9de53d4bf179b1f4e26e114b00f3c22aa0de", "generation_seconds_batch": 21.34575086995028, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.678314208984375, "target_z": 0.7656249378744145, "ce": 5.546053409576416, "genuineness": 4.344170093536377, "identity": 0.6965577602386475, "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_DRAM_reg/84b47dc7488cdf559b5750abdd4ee6971081af343ae3af3ff3dc7d6c1f7955be.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "8571a14cf5b6188576ff049ccd70b44069d0e0f4327dfa36b2d14206fadb795d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "27893cda75c22ae64f6089841b6cee84ff789ba8c4b1397997cca79346121c3d", "generation_seconds_batch": 22.47681180201471, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.767364263534546, "target_z": -0.03393547979367295, "ce": 6.173524379730225, "genuineness": 4.846253395080566, "identity": 0.8720947504043579, "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_DRAM_reg/8571a14cf5b6188576ff049ccd70b44069d0e0f4327dfa36b2d14206fadb795d.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": "85bfd5727e19995db98e8b2c889c792fc81399832ee9862d84afc36f3748f18d", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "958ce5a00cea0408520a55da4ad709e0ca4f8d823cf808a043f5f0f4e25cb7f2", "generation_seconds_batch": 13.52541666093748, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7673230171203613, "target_z": 0.8437498888278997, "ce": 5.657250881195068, "genuineness": 4.2249250411987305, "identity": 0.6917229294776917, "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_DRAM_reg/85bfd5727e19995db98e8b2c889c792fc81399832ee9862d84afc36f3748f18d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "86384a8f5826fd790d91968c3525da4bc6a91e58a400935e70ac37f8bf7adf7b", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "0b5c167f0909e6f46d2139478549a61cad193f1eeb2adaf2a958147639a2da04", "generation_seconds_batch": 20.05379652197007, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9319896697998047, "target_z": 0.9882813619895422, "ce": 5.6532793045043945, "genuineness": 4.18726921081543, "identity": 0.6509497761726379, "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_DRAM_reg/86384a8f5826fd790d91968c3525da4bc6a91e58a400935e70ac37f8bf7adf7b.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "86d04b31b61ed754aa4cef67cdad59434e5e2ae9d49f994ac3a0cafcde52be66", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "7557abdf50dbe19a84c3d63052428e1adf55809ad29fee74ed60c4f259e71470", "generation_seconds_batch": 22.577014012029395, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.947329044342041, "target_z": 0.12402349042936392, "ce": 6.3641486167907715, "genuineness": 4.457139015197754, "identity": 0.5804511308670044, "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_DRAM_reg/86d04b31b61ed754aa4cef67cdad59434e5e2ae9d49f994ac3a0cafcde52be66.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "86d6e0a8c341ec171908b51c45730029eb45506ace0ff1cfe749d55500ce8807", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "9c604d31c89f6e55be397396c76fc5b1d05600c9c267f748606e3ea6bafd8d9c", "generation_seconds_batch": 15.992547197965905, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.069953441619873, "target_z": 1.1093751406000092, "ce": 5.7247633934021, "genuineness": 4.406930446624756, "identity": 0.7546420097351074, "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_DRAM_reg/86d6e0a8c341ec171908b51c45730029eb45506ace0ff1cfe749d55500ce8807.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "871aacb4a4dbc949c63eb9729159ea705300f114896538a44afb0597dd2dbd0b", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "cca23aa0a9208268d8bb4ac7b351a4af12478ac27d30c7d3f638fd8d39e6bf7d", "generation_seconds_batch": 21.788887670030817, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3965866565704346, "target_z": -0.3593749836511617, "ce": 6.372091293334961, "genuineness": 4.319066047668457, "identity": 0.5657392740249634, "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_DRAM_reg/871aacb4a4dbc949c63eb9729159ea705300f114896538a44afb0597dd2dbd0b.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "8742f825032dd9c2cf09b38b88071e51b5ea104e2c706df0c5602434258c189c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "d0b586a3dff9baa5977c1c7a6db89361a830f8c4e0bdbbac38c0103dda76ac9b", "generation_seconds_batch": 20.354645093902946, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6369106769561768, "target_z": -0.14843740681162182, "ce": 6.419747352600098, "genuineness": 4.281409740447998, "identity": 0.4257830083370209, "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_DRAM_reg/8742f825032dd9c2cf09b38b88071e51b5ea104e2c706df0c5602434258c189c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "8762a603de412db2b9e06f66f6143e2b61a508d084753c859555f9f21237b8ce", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "657c0b1db458afb46fc363212caacdeb9a1c15526c7559092e3a8a6a4d7df5b5", "generation_seconds_batch": 22.153786478098482, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.998746156692505, "target_z": 1.046874970572091, "ce": 5.855817794799805, "genuineness": 4.394378662109375, "identity": 0.7236518263816833, "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": 0.03999999999999915, "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_DRAM_reg/8762a603de412db2b9e06f66f6143e2b61a508d084753c859555f9f21237b8ce.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "87708bbb3905f63efe7b951b27bd45c77c53cbfe1607c1520567db9936386c97", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "b37421eca0397ea1580b900d476bb91e72a8f4376377379645ef2dc100547414", "generation_seconds_batch": 21.26309336000122, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.531449794769287, "target_z": 0.636718873433729, "ce": 5.593709468841553, "genuineness": 4.432034492492676, "identity": 0.4640570282936096, "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_DRAM_reg/87708bbb3905f63efe7b951b27bd45c77c53cbfe1607c1520567db9936386c97.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "87b64be910f2cd337ec941c24c5308a5658d398d7c8b5871dbae9af6096f15e7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "de803608e935f49f79cbcfb964850994fed3792bf6edf094324e88d0eb9fc5a8", "generation_seconds_batch": 16.027165808947757, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1843152046203613, "target_z": 0.3320313554500069, "ce": 6.0146708488464355, "genuineness": 4.7332844734191895, "identity": 0.8298500180244446, "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_DRAM_reg/87b64be910f2cd337ec941c24c5308a5658d398d7c8b5871dbae9af6096f15e7.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "881d504df5e56964cc1d3b1d4b5ec67947ffc9b0883067321ef46973eb4c75b5", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "5ffdb98632069cb357650ddd49bbee1ea2fda4cabd27d8289133a47a039e7e0b", "generation_seconds_batch": 20.50956071401015, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.357882499694824, "target_z": 0.4843751144418679, "ce": 6.125868320465088, "genuineness": 4.419482707977295, "identity": 0.6314733624458313, "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": 0.03999999999999915, "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_DRAM_reg/881d504df5e56964cc1d3b1d4b5ec67947ffc9b0883067321ef46973eb4c75b5.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "889c272d892696a9ca5006179f6b5b43331e65569b858f3e4d2dd51d75918a57", "bucket": "vn_S_DRAM_reg", "phase": "wording", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "e76a52f43677070b4161bd81742a7d8e00c7fd044c3f4735b11cbc374ead9760", "generation_seconds_batch": 18.50288703397382, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0074100494384766, "target_z": 0.17675788463924888, "ce": 5.871703147888184, "genuineness": 4.256305694580078, "identity": 0.49358806014060974, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/889c272d892696a9ca5006179f6b5b43331e65569b858f3e4d2dd51d75918a57.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "88ed84cd237aaf4034abefaed4db8aa209c423885085ac472ba18dd5888b7e28", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "80d489b55d7b0e8a72f6439aac30e532abb35ffeae16d745496999c01670aea4", "generation_seconds_batch": 20.274366572964936, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.898374080657959, "target_z": 0.0810546627723821, "ce": 6.181467056274414, "genuineness": 4.6705241203308105, "identity": 0.6074042916297913, "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_DRAM_reg/88ed84cd237aaf4034abefaed4db8aa209c423885085ac472ba18dd5888b7e28.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "89008de22be58b8eb511b8dc0dee9e255c2c2dd516635d82a96083bc936e90e6", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "c4eb14d510ec582a00fb33466ac758159e53607f507850b1ea1f84f15d76f121", "generation_seconds_batch": 15.768428527982906, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.504746913909912, "target_z": 0.6132811788825535, "ce": 6.1894097328186035, "genuineness": 4.482243061065674, "identity": 0.6042971611022949, "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": 0.03999999999999915, "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_DRAM_reg/89008de22be58b8eb511b8dc0dee9e255c2c2dd516635d82a96083bc936e90e6.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "8964fdfe3f836cc54cb30ba1caffd370d779b98db5ddbf64fe982e5b617f0d44", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "0e50340aa78ae0ca4f56afd21249dc5e17b7abc7a816ec505e48f37ffae2617f", "generation_seconds_batch": 14.719412734964862, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.972043514251709, "target_z": 1.0234374852860455, "ce": 5.8875885009765625, "genuineness": 4.419482707977295, "identity": 0.2749914526939392, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/8964fdfe3f836cc54cb30ba1caffd370d779b98db5ddbf64fe982e5b617f0d44.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "89f0b82a899b296eb3264e39f15fe4c3791e2d1841275ed87796c7c747a3597e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "adcece8f99f8cfd5b83f820a0262cc540526a78e31260fa440776b844218c79a", "generation_seconds_batch": 24.13657992298249, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.651611566543579, "target_z": 0.742187452588369, "ce": 6.117925643920898, "genuineness": 4.15588903427124, "identity": 0.5710487961769104, "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": 0.03999999999999915, "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_DRAM_reg/89f0b82a899b296eb3264e39f15fe4c3791e2d1841275ed87796c7c747a3597e.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": "8a785e55ca09888c4f67b7a584e40c5ca52bdb292d662e1db2c37b914f04413c", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "37d28e8382779c9dfa88aa8f276a3659313d52ce27dd5473f2fad0589cb25fe7", "generation_seconds_batch": 21.146918909973465, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.345880508422852, "target_z": 1.3515622792906832, "ce": 6.332377910614014, "genuineness": 4.231201171875, "identity": 0.49946844577789307, "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_DRAM_reg/8a785e55ca09888c4f67b7a584e40c5ca52bdb292d662e1db2c37b914f04413c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "8a8e1464b2465219234c667dfde5fa4636045ab361d6f8e6e119c5fa055b9bfb", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "c74e2dacadf7f29d2d6dd3fc83850b9ec94f04072cd2c9b83d99cbc33402d43c", "generation_seconds_batch": 15.081610720953904, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.041900873184204, "target_z": 0.20703122465930068, "ce": 6.1894097328186035, "genuineness": 4.6705241203308105, "identity": 0.3815903663635254, "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_DRAM_reg/8a8e1464b2465219234c667dfde5fa4636045ab361d6f8e6e119c5fa055b9bfb.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "8ac145695a30d6a3b083b66516eb61c76e33e7e477511a2a5bdb927c1f365b73", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "9bff4dc39ca844ffa8f99a7b540d67304b399afcd79d94d98c545f78b6db018e", "generation_seconds_batch": 15.824774271110073, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.358757972717285, "target_z": -0.39257803549011044, "ce": 6.125868320465088, "genuineness": 4.65797233581543, "identity": 0.6781672239303589, "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_DRAM_reg/8ac145695a30d6a3b083b66516eb61c76e33e7e477511a2a5bdb927c1f365b73.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "8ba50aed61e3905cc827baf14032f9adf9ab59d546457de309fc908b2948eaae", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "a2b1ecbbdc19d9a7d3783c18029693dc7b11856b51d9b22d85158e622054e4c8", "generation_seconds_batch": 14.483692134032026, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.265772819519043, "target_z": 1.2812500326976766, "ce": 5.871703147888184, "genuineness": 4.344170093536377, "identity": 0.6106903553009033, "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_DRAM_reg/8ba50aed61e3905cc827baf14032f9adf9ab59d546457de309fc908b2948eaae.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "8c0320156ebd670453b7b72d3a6bd2b45b31edf23bfcab234e4bb89c23cd1ef5", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "4760533a8fcafeefeadb13f3f6ba2f2ec59758364bf14b2f24f1a50fc164d8a1", "generation_seconds_batch": 20.708036763011478, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.604006290435791, "target_z": 1.5781248463209203, "ce": 5.998785495758057, "genuineness": 4.13078498840332, "identity": 0.4395084083080292, "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_DRAM_reg/8c0320156ebd670453b7b72d3a6bd2b45b31edf23bfcab234e4bb89c23cd1ef5.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "8c298abae90bd610ad2b9acfd2f00930c69bd2023328d11631586f3e6c1a2a07", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "d5637c2c7139ab2ea973984391997f296ca6066e9b4f326e4cee7700e57bc4e9", "generation_seconds_batch": 15.27937819401268, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1353602409362793, "target_z": 0.28906252779302505, "ce": 6.157639026641846, "genuineness": 4.808597087860107, "identity": 0.46581780910491943, "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_DRAM_reg/8c298abae90bd610ad2b9acfd2f00930c69bd2023328d11631586f3e6c1a2a07.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "8cb84cee6fbc35e04573277c999a385e643da5ef80fdfc5c40a0895d3ce98557", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "92a02887cbd71a2a1af02eff16c1445ac192866437dc133f0169743aa59e0eb8", "generation_seconds_batch": 21.13125059194863, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.47939395904541, "target_z": 1.468749914986041, "ce": 5.772419452667236, "genuineness": 4.042920112609863, "identity": 0.19732427597045898, "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_DRAM_reg/8cb84cee6fbc35e04573277c999a385e643da5ef80fdfc5c40a0895d3ce98557.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "8cde23d40c0e0bd6b1bd8cb748a6dcffeb17bd36a097dde1f0e0a102cc7b2305", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "3a6223f5314efc9b849eddececa648e52f0361e227621066cf83a650f6983c25", "generation_seconds_batch": 22.19174836599268, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.918638229370117, "target_z": 0.9765625147139545, "ce": 6.292664527893066, "genuineness": 4.457139015197754, "identity": 0.6602601408958435, "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_DRAM_reg/8cde23d40c0e0bd6b1bd8cb748a6dcffeb17bd36a097dde1f0e0a102cc7b2305.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": "8d07763aaec4d5c90bb7678e6443239a049417b77c902f01475bda9aa0db3fad", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "464643a513c9ce33731086be3afe76dfae4fa430eebe9ffdb6361b252514a3a0", "generation_seconds_batch": 15.75897237402387, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9364399909973145, "target_z": 0.9921875049046515, "ce": 5.657250881195068, "genuineness": 4.080576419830322, "identity": 0.37310928106307983, "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_DRAM_reg/8d07763aaec4d5c90bb7678e6443239a049417b77c902f01475bda9aa0db3fad.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "8d7fc0c8a4bb52b21a27d0ecb70e5400894a0cc2b9bb1699f987ec6a937985ee", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "f1bfc4714f0ea4230832445698ef53179c2a418470dad567ec046330debc37d8", "generation_seconds_batch": 20.73877645994071, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.132259368896484, "target_z": 1.164062397002319, "ce": 5.677107334136963, "genuineness": 4.218649387359619, "identity": 0.21820449829101562, "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": 0.03999999999999915, "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_DRAM_reg/8d7fc0c8a4bb52b21a27d0ecb70e5400894a0cc2b9bb1699f987ec6a937985ee.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "8e0558028593ee2ba9fd5cd4577e1937ddc0edf876bc92b922072432e95de797", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "6e0e5ce383787cc7a2f4072a2576c7fe81386d35352ca7b2923e4b84ef6e0ba5", "generation_seconds_batch": 21.057567595969886, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.124234199523926, "target_z": 0.2792969612401219, "ce": 6.09409761428833, "genuineness": 4.545003414154053, "identity": 0.7148035764694214, "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_DRAM_reg/8e0558028593ee2ba9fd5cd4577e1937ddc0edf876bc92b922072432e95de797.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "8e24b787bf7e6918cd4ebae55d123a09a2fc79bba2518bafc15e72569f59cd32", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "03825696216c8c0a944a3f9dc74c1ffa1c41f186e8e52ed8bc89ec5ee9063483", "generation_seconds_batch": 21.440973090939224, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.336505651473999, "target_z": -0.4121093778610467, "ce": 6.38003396987915, "genuineness": 4.319066047668457, "identity": 0.6189820170402527, "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_DRAM_reg/8e24b787bf7e6918cd4ebae55d123a09a2fc79bba2518bafc15e72569f59cd32.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "8eeb41c1146f70e55098af296495374a5a4fe36a81e8a62f972624fba7430a2d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "be06f22fea4142d71d2505d8722355ce28194a39837df561ba71ea2a2d5db166", "generation_seconds_batch": 21.348820040933788, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.158962249755859, "target_z": 1.1875000915534943, "ce": 5.180690288543701, "genuineness": 4.017816066741943, "identity": 0.3116794526576996, "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": 0.03999999999999915, "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_DRAM_reg/8eeb41c1146f70e55098af296495374a5a4fe36a81e8a62f972624fba7430a2d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "9001dc675e3c90335a5b64238dd98ec3b1854952664268fb81c514c31c570026", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "0a4f121b1cc8d1b4fa5e91903af37450d3d02a3b46156dd8c1a17cbf70323ec5", "generation_seconds_batch": 15.315205003018491, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7370455265045166, "target_z": -0.06054688930523349, "ce": 6.0305562019348145, "genuineness": 4.457139015197754, "identity": 0.5141714215278625, "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_DRAM_reg/9001dc675e3c90335a5b64238dd98ec3b1854952664268fb81c514c31c570026.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "90059c33caa26ea288712bac7833dccf6d3e30d81a24c923d9498cb9debeb788", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "f4d7c29542e8f4d5d1cfdc38c6a41bd0b09ecd699e4eead6da4acdcab37a27f8", "generation_seconds_batch": 14.773813166073523, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5670533180236816, "target_z": 0.667968853815123, "ce": 6.284721851348877, "genuineness": 4.325342178344727, "identity": 0.33986708521842957, "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_DRAM_reg/90059c33caa26ea288712bac7833dccf6d3e30d81a24c923d9498cb9debeb788.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "909acf30d2c4a8b1ef829a9330adfdb23901c529dcdd57f395d7011d43377ebd", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "bcafc6f6991320cf5dcb60ed79561148db75865e694b6d00b89585f41881ae43", "generation_seconds_batch": 15.392445442965254, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.36098313331604, "target_z": -0.3906249640325558, "ce": 6.340320587158203, "genuineness": 4.068024158477783, "identity": 0.5658299922943115, "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_DRAM_reg/909acf30d2c4a8b1ef829a9330adfdb23901c529dcdd57f395d7011d43377ebd.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "90f3fbbbd2409fada8d6dc9761235b36f9ba4534180b0e5761d5e426974a97ef", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "839d6bd0bbb58e0325d76dabe32756755d583d09011f07928d7b33bd4eea4445", "generation_seconds_batch": 21.440965233021416, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.354781627655029, "target_z": 1.3593749836511617, "ce": 5.919358730316162, "genuineness": 4.143336772918701, "identity": 0.392644464969635, "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_DRAM_reg/90f3fbbbd2409fada8d6dc9761235b36f9ba4534180b0e5761d5e426974a97ef.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "911cc66694c5164b7caca0785c4b51fd3a634cea8bf18ca5372301873ce8b71d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "e67f50194b6830a5354c76709598d79fd628015ac1841327f429e7c74698b409", "generation_seconds_batch": 15.392445442965254, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2319202423095703, "target_z": -0.5039062475476742, "ce": 6.38003396987915, "genuineness": 4.118232727050781, "identity": 0.5740347504615784, "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_DRAM_reg/911cc66694c5164b7caca0785c4b51fd3a634cea8bf18ca5372301873ce8b71d.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": "923e308d09b0cedf2507a0052014f76f330c54f78280f29046020e8b4105f13a", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I wrote the number on the envelope.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "16c5cbee2c20d6c6a5ae87ad3def3f3d82e98d85bd33f5610db88bfa26e88581", "generation_seconds_batch": 23.220257638953626, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.802926540374756, "target_z": 0.8749998692092938, "ce": 6.078212261199951, "genuineness": 3.911123275756836, "identity": 0.33788231015205383, "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_DRAM_reg/923e308d09b0cedf2507a0052014f76f330c54f78280f29046020e8b4105f13a.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "92724963e21399a22591446d84f2f8dac0e08a947e560baf4fb21079b0e7b9f9", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "df8ea09e909cac13a4df0a785da90150c5d59022c77dc06dc70b7a1f392456f1", "generation_seconds_batch": 20.66200865409337, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.114457607269287, "target_z": 1.1484374068116219, "ce": 6.109982967376709, "genuineness": 4.262581825256348, "identity": 0.5599343776702881, "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_DRAM_reg/92724963e21399a22591446d84f2f8dac0e08a947e560baf4fb21079b0e7b9f9.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "9319f3fae0ca460c34e759169356ea7ab2d5c605406e559d36bcf5ec1e30d09d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "333a1ac6f72ddc276c663ddf25e1c7a5b0413547192d828e8f32eb33f1932b1e", "generation_seconds_batch": 20.758678595069796, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.337855339050293, "target_z": 0.4667968435284863, "ce": 6.213237762451172, "genuineness": 3.9425034523010254, "identity": 0.2858242690563202, "hypothesis": "The door is still open. I'd put your coat on the chair. Tell me when you are ready.", "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_DRAM_reg/9319f3fae0ca460c34e759169356ea7ab2d5c605406e559d36bcf5ec1e30d09d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "9324662f7fa6e066e98535b090cc13e6a0c680922676c64f363659c8bf60bc5d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "93a0e8e8c03752734fe271ded9d4673ca2ad11f56253c64cc32cd7622b1557d9", "generation_seconds_batch": 21.171578763984144, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.758422374725342, "target_z": 0.8359376029976812, "ce": 5.8875885009765625, "genuineness": 4.143336772918701, "identity": 0.5966484546661377, "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_DRAM_reg/9324662f7fa6e066e98535b090cc13e6a0c680922676c64f363659c8bf60bc5d.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": "934bff232b2095ae77fbb7682d01859b530b31e44baae057ffccd3eca4c77660", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "772e43e797edd81a4cb5ff391f241f4dc6f73fab76d4f952a9c521fbe7dcfed0", "generation_seconds_batch": 21.321423097047955, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.847431182861328, "target_z": 0.9140625539511663, "ce": 5.276002407073975, "genuineness": 3.936227321624756, "identity": 0.5051881074905396, "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_DRAM_reg/934bff232b2095ae77fbb7682d01859b530b31e44baae057ffccd3eca4c77660.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "940fdc854aedea63e4d968f11314ed55f53211af1391a93d2f0dd6abdff24253", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "c5741b1f5f987b69c21ecf168f50b26fbf1eb0ddbf455eb7aef1de7f84db27c8", "generation_seconds_batch": 22.04210061696358, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.945340871810913, "target_z": 1.0, "ce": 5.998785495758057, "genuineness": 4.319066047668457, "identity": 0.5401613712310791, "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": 0.03999999999999915, "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_DRAM_reg/940fdc854aedea63e4d968f11314ed55f53211af1391a93d2f0dd6abdff24253.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "9474207df01b18e1bc803d0346aaedba5e4875e1902d1e1e44f573dc181ade33", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "5b9521a3b477352c0340efa8f7a83d57b65898878428854c0ff31b3255aabae1", "generation_seconds_batch": 20.61788582499139, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7406206130981445, "target_z": 0.8203126128069841, "ce": 5.871703147888184, "genuineness": 4.444586753845215, "identity": 0.7206200361251831, "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_DRAM_reg/9474207df01b18e1bc803d0346aaedba5e4875e1902d1e1e44f573dc181ade33.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "94765005732d8ccbf74e981d4028821cfa595f28e407bcb1a11c37b09294f5a4", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "2f3d54821f8905df838a29b8e8b2142c7ca69e69c39a44b8f852250a84598fc2", "generation_seconds_batch": 17.90900560596492, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2466213703155518, "target_z": 0.3867188211174465, "ce": 6.1020402908325195, "genuineness": 4.507347106933594, "identity": 0.4788750112056732, "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_DRAM_reg/94765005732d8ccbf74e981d4028821cfa595f28e407bcb1a11c37b09294f5a4.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "94dc243c280e6844ee18e9e7647278988159deadc262e3b8267982fe8bd80b46", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "9660bec828cae7158e25f6cfd09d10caf2c592be7ffa7d57655e7a2555bf8a79", "generation_seconds_batch": 19.773692814982496, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.212367534637451, "target_z": 1.2343750621255853, "ce": 5.462655067443848, "genuineness": 3.973883628845215, "identity": 0.35933399200439453, "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_DRAM_reg/94dc243c280e6844ee18e9e7647278988159deadc262e3b8267982fe8bd80b46.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "952df21b324542bae40acb8120c8e2eb3741e3f76962c596632c3278f3e6d87d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "1b992bf1690bd38de7a0cc5c6aee68da49cbd4a6dcf0f7fb25d5a5982a17e04f", "generation_seconds_batch": 21.18548683600966, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3000266551971436, "target_z": 0.4335937916895376, "ce": 5.9431867599487305, "genuineness": 4.482243061065674, "identity": 0.7898187637329102, "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_DRAM_reg/952df21b324542bae40acb8120c8e2eb3741e3f76962c596632c3278f3e6d87d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "954eda0d471d7f671ea2316a89eb274f29fcb7b908074b42d0cae0b78a4b2540", "bucket": "vn_S_DRAM_reg", "phase": "wording", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "927cd1cfa5bbb522610658a76f5b17fda14f87ef5e4974d482a850929bc49c69", "generation_seconds_batch": 14.078537620953284, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9542417526245117, "target_z": 1.0078124950953484, "ce": 6.308549880981445, "genuineness": 3.9927117824554443, "identity": 0.3403468132019043, "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_DRAM_reg/954eda0d471d7f671ea2316a89eb274f29fcb7b908074b42d0cae0b78a4b2540.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "96450ef22d9e14d7f5be6390a0b91a241e216c26762ac5c776d7ac02d4f79dd0", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "9a3f813c3998cd908804453de96e1afcbc85fa6c3682e23fd13efaf64d544c94", "generation_seconds_batch": 21.02504669909831, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.176764011383057, "target_z": 1.2031250817441914, "ce": 5.998785495758057, "genuineness": 3.9173994064331055, "identity": 0.4086894392967224, "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/manual-comparison-v1/audio/vn_S_DRAM_reg/96450ef22d9e14d7f5be6390a0b91a241e216c26762ac5c776d7ac02d4f79dd0.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "9686888e717705a8ef31cb21820ff05e82bb6685ee285e36429f1a8ecf0c3b39", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "437a89992bee269a31e178da13d229b973a5b4265cb29117e37fe4410f66326a", "generation_seconds_batch": 20.3966804939555, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.776543140411377, "target_z": -0.025878981556836305, "ce": 6.284721851348877, "genuineness": 4.557555198669434, "identity": 0.6138471364974976, "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_DRAM_reg/9686888e717705a8ef31cb21820ff05e82bb6685ee285e36429f1a8ecf0c3b39.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": "970de662baaae57e514ab25473574151544775cba9f80918511b89330357386d", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "6b3949bcd10cbded5538fa8b16b45f73f9e4c1d8a1a2693d3dfd81d799d9caa1", "generation_seconds_batch": 14.893597115995362, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.488295078277588, "target_z": 1.4765626193465193, "ce": 5.458683967590332, "genuineness": 3.6977379322052, "identity": 0.7373645901679993, "hypothesis": "J'ai écrit le numéro sur l'enveloppe, la bien et vérifié les deux derniers chiffres. Appelle-moi quand tu auras un moment.", "wer": 0.13636363636363635, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_DRAM_reg/970de662baaae57e514ab25473574151544775cba9f80918511b89330357386d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "971866b1464ce628558bba40a2314c36b82060f9bb96ec89aad52eeeddcd8d8a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "026be2f74884d6d28c66a37a41002060fd44f4ec0c2158860128714e3d2f854c", "generation_seconds_batch": 20.972861030953936, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1042070388793945, "target_z": 0.2617186903267403, "ce": 6.260893821716309, "genuineness": 4.319066047668457, "identity": 0.6052130460739136, "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_DRAM_reg/971866b1464ce628558bba40a2314c36b82060f9bb96ec89aad52eeeddcd8d8a.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "985d2f03461fea276e6a69bc3d45c35cfb36969ba2a5387f66905ee2868b9cf4", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "a605ac6926e4e752b13999a4d405bbb84a3cbf1edb3b857297ca59b17f0d6d54", "generation_seconds_batch": 16.06255209504161, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.705016851425171, "target_z": 0.7890624231604602, "ce": 6.38003396987915, "genuineness": 4.469690799713135, "identity": 0.5287193059921265, "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_DRAM_reg/985d2f03461fea276e6a69bc3d45c35cfb36969ba2a5387f66905ee2868b9cf4.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "98fddbec327dbb18e1777e21ba4b5bb5f7b30687eb88c1c7d095be97829ac504", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "53c9e419f6797b60418a39b5080300421a7d8ac8b1a9a6fbcd2773413323dd11", "generation_seconds_batch": 15.27937819401268, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6872150897979736, "target_z": 0.7734374329697631, "ce": 5.621508598327637, "genuineness": 4.519899368286133, "identity": 0.3840077519416809, "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_DRAM_reg/98fddbec327dbb18e1777e21ba4b5bb5f7b30687eb88c1c7d095be97829ac504.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "99096d594fdcdbe6a98a99ebf4327d01a91d528bae0ed51edc2daedc592c416f", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "0bcfc5a6835012471752540b196033b064f5eb8046c1bc13452c4643ce424efe", "generation_seconds_batch": 22.183748372015543, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8905858993530273, "target_z": 0.07421880803837588, "ce": 6.062326908111572, "genuineness": 3.8671910762786865, "identity": 0.170360267162323, "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_DRAM_reg/99096d594fdcdbe6a98a99ebf4327d01a91d528bae0ed51edc2daedc592c416f.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "991ac45c9b46e700595d1fecf3321d30d4de8d4c0d0234ab59823fcf9dfcaba7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "555646b540b45ce2b171e47b2636bd0d3bcb2b5c3b6e5b44ae86096421859eb6", "generation_seconds_batch": 15.974917615996674, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.399286270141602, "target_z": 1.3984376683930342, "ce": 5.270045757293701, "genuineness": 3.8295347690582275, "identity": 0.5094864368438721, "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_DRAM_reg/991ac45c9b46e700595d1fecf3321d30d4de8d4c0d0234ab59823fcf9dfcaba7.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": "99525237f9e9da632743d152c018fd86a57c6489b485abfb72c0efa7ec9282c9", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c903a4714aae9aad1c5aae482a905b64c0da65fc5fce522d52d5d76a150aafb4", "generation_seconds_batch": 21.42373379005585, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.043250560760498, "target_z": 1.0859374460488338, "ce": 5.569881439208984, "genuineness": 4.137060642242432, "identity": 0.7835960388183594, "hypothesis": "La route tourne à gauche après le pont. Continue jusqu'à la petite barrière en bois. Je je t'attendrai de l'autre côté.", "wer": 0.05, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_DRAM_reg/99525237f9e9da632743d152c018fd86a57c6489b485abfb72c0efa7ec9282c9.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "99750e62b972395746ac16660c13382f41f1f5968d51a9bf7249f7ab3b9b37c2", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "717adc4ecbafd211f0e6724d72a16f2c461e157c2a75456f50e922debcdce794", "generation_seconds_batch": 20.95962341793347, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0274369716644287, "target_z": 0.19433594628750056, "ce": 6.23706579208374, "genuineness": 4.65797233581543, "identity": 0.5900728702545166, "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_DRAM_reg/99750e62b972395746ac16660c13382f41f1f5968d51a9bf7249f7ab3b9b37c2.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "997dcbe9d77d43bea3475eb34e2c9a6bb42c691c944289bd2406776877740228", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "6ade309a8a010c2b74292c94e690f3e58e7203d803065db889b06b67f9998b2a", "generation_seconds_batch": 21.19263188203331, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.176764011383057, "target_z": 1.2031250817441914, "ce": 5.242246150970459, "genuineness": 4.118232727050781, "identity": 0.4107770323753357, "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_DRAM_reg/997dcbe9d77d43bea3475eb34e2c9a6bb42c691c944289bd2406776877740228.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "002bb986258b9fa6cb75d8d3e31a6ec8f449fb80416f42c82646c594058ecf25", "generation_seconds_batch": 15.506583926035091, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.367658853530884, "target_z": -0.3847655403947619, "ce": 6.316492557525635, "genuineness": 4.557555198669434, "identity": 0.5572497844696045, "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_DRAM_reg/99bf1b7d3eed7b885fdae888ab00b4960791c3971d259b46f771306d2801e39b.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "9a266da673dc5120f6a9c137c8dfaff1d52f7a822b289759c4a2006c7e0ff533", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "34dc151e7142178f85d9f12f40f5faad1daf08d4ebb577389fd6ddbaf1d40cfa", "generation_seconds_batch": 20.61641469306778, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.976256847381592, "target_z": 0.1494140471729641, "ce": 5.935244083404541, "genuineness": 4.507347106933594, "identity": 0.8426653742790222, "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_DRAM_reg/9a266da673dc5120f6a9c137c8dfaff1d52f7a822b289759c4a2006c7e0ff533.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": "9a2f804e1a4f112a3a0fddad5f3a9a194c46877440fbfcb76b2779fce58ebc1b", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "9dadbb30db4ac80ebb2feb01de28041967ba415cfb4b186664c44b77d26dfecc", "generation_seconds_batch": 14.658118849969469, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.998746156692505, "target_z": 1.046874970572091, "ce": 5.248203277587891, "genuineness": 4.13078498840332, "identity": 0.39405977725982666, "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_DRAM_reg/9a2f804e1a4f112a3a0fddad5f3a9a194c46877440fbfcb76b2779fce58ebc1b.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "9a44cb985331091b2a31ac8b0fe398784ccd5c5867f2d279926cc04caa48a1ca", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "b68c5e3ceee861a6b3b4cb874271f82773df3db07393fe0351f080407f4bb1ce", "generation_seconds_batch": 16.00596959097311, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0361008644104004, "target_z": -0.6757811396453416, "ce": 6.252951145172119, "genuineness": 4.7458367347717285, "identity": 0.5493936538696289, "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_DRAM_reg/9a44cb985331091b2a31ac8b0fe398784ccd5c5867f2d279926cc04caa48a1ca.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "9a56f1b336410def9409b7ca9fad3cad23fa9521145f01967f5a3c790f87325e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "f3d87dfb429fc9d687d372b0faf1e79c15de1bcb347e7b90c38b849c3f67f940", "generation_seconds_batch": 22.0043279719539, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.235495090484619, "target_z": 0.3769530452994134, "ce": 6.109982967376709, "genuineness": 4.6956281661987305, "identity": 0.6642848253250122, "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_DRAM_reg/9a56f1b336410def9409b7ca9fad3cad23fa9521145f01967f5a3c790f87325e.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "9a5897464efafb86c9e9122760ecab77c46a79aa301d7ae0677923f606cf231a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "e4054925a4cdb6594b1ddd156f87bfa732a4085086f28170bbf2dc8e763c52ea", "generation_seconds_batch": 22.35787030204665, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.656937599182129, "target_z": -0.13085934516337014, "ce": 6.443574905395508, "genuineness": 4.545003414154053, "identity": 0.4065096974372864, "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_DRAM_reg/9a5897464efafb86c9e9122760ecab77c46a79aa301d7ae0677923f606cf231a.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "9b3f9be0d4d063396354185bc17eb15420006909de2e0e51be44fb3c71580350", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "6c5d11acdfbf310317d7ba17412af1658de857df359105989908e97735376b6c", "generation_seconds_batch": 20.96969827800058, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9818198680877686, "target_z": 0.1542968304494157, "ce": 6.284721851348877, "genuineness": 4.607763767242432, "identity": 0.7030944228172302, "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_DRAM_reg/9b3f9be0d4d063396354185bc17eb15420006909de2e0e51be44fb3c71580350.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "9b98ed0344b3fdfdac6c19dc0efe23a99e21861010d30b0a3b14f2d20f052770", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "1b33a189c81921c27823dfa29e4ecfd85cd2d82f9ad150662ee5b1e138389b1d", "generation_seconds_batch": 20.621511646080762, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7914938926696777, "target_z": -0.012756383789737095, "ce": 6.173524379730225, "genuineness": 3.9425034523010254, "identity": 0.3197195529937744, "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/manual-comparison-v1/audio/vn_S_DRAM_reg/9b98ed0344b3fdfdac6c19dc0efe23a99e21861010d30b0a3b14f2d20f052770.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "9c06f5195b8fa3800c855b0c977e2cba358fc72f4cbb73df28aaed5067407c36", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "5f82963d3978ff65b71813ec1dd5ac435acabfd32f421732db90258ab1f3540d", "generation_seconds_batch": 21.031203671940602, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.071941375732422, "target_z": 0.23339842176424316, "ce": 6.165581703186035, "genuineness": 4.557555198669434, "identity": 0.7524535059928894, "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_DRAM_reg/9c06f5195b8fa3800c855b0c977e2cba358fc72f4cbb73df28aaed5067407c36.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "870f9503ff8842cbddeb7a5a44e3a05090b251f69b29219bd06d2e6503b7fdd6", "generation_seconds_batch": 15.767950225970708, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9159388542175293, "target_z": -0.7812499280651116, "ce": 6.324435234069824, "genuineness": 4.381826400756836, "identity": 0.6593779921531677, "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_DRAM_reg/9c09e940c5dab53287ffae2d1a8fa5296f41b2df71714656aaa2bb8830bf5162.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "9c0c72ad20f5dcd90671724118a46bcaeea110dbeb485e785a330069e26c50a0", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "2ab54f64c36041a33819ca2bc087698a069f3020ab4b81fef1be6491e1f1d82e", "generation_seconds_batch": 22.101238091941923, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9128379821777344, "target_z": 0.09374994114418221, "ce": 6.109982967376709, "genuineness": 4.457139015197754, "identity": 0.6907817125320435, "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_DRAM_reg/9c0c72ad20f5dcd90671724118a46bcaeea110dbeb485e785a330069e26c50a0.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "9c921eabae32251e74aaae044d7959edc865b789c5d5274523db7bd3e7cc33fc", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "5ac6ee232dec8329fb1b63e0c2d5252b973884c30dcddd6f4e79777365ea7516", "generation_seconds_batch": 20.45422521501314, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.38903546333313, "target_z": 0.5117187426430228, "ce": 5.6413655281066895, "genuineness": 4.306513786315918, "identity": 0.7408486604690552, "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_DRAM_reg/9c921eabae32251e74aaae044d7959edc865b789c5d5274523db7bd3e7cc33fc.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "e204f38585567ebd434868bc5a138a8046f1d4caab2b9be3b415e96be3e40a72", "generation_seconds_batch": 15.754555492079817, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5946314334869385, "target_z": -0.18554681083080976, "ce": 6.300607204437256, "genuineness": 4.18099308013916, "identity": 0.5307666063308716, "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_DRAM_reg/9c969db20a6db380e08592f6aed9c9189b3db634d677ebe8f573f593ebdfd2d3.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "9d7bede279ef9c7c31ca410b180f26b22a5d440fc755257f03c29440baf63603", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "75fd9ef32e59c07c4f30ab20f8a4440cc936221af35d59e8e58bcfe699225c2c", "generation_seconds_batch": 21.78752606897615, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7592978477478027, "target_z": -0.04101554693429724, "ce": 6.141753673553467, "genuineness": 4.406930446624756, "identity": 0.422296941280365, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/9d7bede279ef9c7c31ca410b180f26b22a5d440fc755257f03c29440baf63603.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "9de7e23be275c2da778f7823232bfc11bf3549a963310cd1ae38628e56baa285", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "04d8b3d382a5109e276d5f68a85ff966e9d37983f504d1b1648c21e7e79f73b0", "generation_seconds_batch": 20.431517314980738, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.481145143508911, "target_z": -0.28515617561278583, "ce": 6.23706579208374, "genuineness": 4.344170093536377, "identity": 0.6505029201507568, "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_DRAM_reg/9de7e23be275c2da778f7823232bfc11bf3549a963310cd1ae38628e56baa285.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "9eb58f1b49ed6ac8c8cb1a8fd07b485dc1f03a515a032acd2f781a49a4ec4c92", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "033d7fd72f93cd7200320d5684d31f399dfa4ba161cc165cc78ab4404ca2fd6a", "generation_seconds_batch": 21.772093457984738, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1754140853881836, "target_z": 0.3242186510895284, "ce": 6.173524379730225, "genuineness": 4.356721878051758, "identity": 0.4956561028957367, "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": 0.03999999999999915, "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_DRAM_reg/9eb58f1b49ed6ac8c8cb1a8fd07b485dc1f03a515a032acd2f781a49a4ec4c92.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "9eb769a11b0a247f75b5cf9c09acbebb174027f1f6023c43ad29069f9afbacf7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "890b783639a05407f3d6c466dadd72a766f53e8439037cfff1b2a74ee9f2e4c4", "generation_seconds_batch": 15.274529071059078, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.381010055541992, "target_z": -0.3730469023843041, "ce": 6.435632705688477, "genuineness": 4.6203155517578125, "identity": 0.6140930652618408, "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_DRAM_reg/9eb769a11b0a247f75b5cf9c09acbebb174027f1f6023c43ad29069f9afbacf7.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "9f199732b39d0ba7f74fed87b5918673a2d951f07714fe1b0deeb9f621a0a6a4", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "d9ae3e525b1b4686e858309b688eca77b568c2f2c4d591d31b3c45f8b3333f6b", "generation_seconds_batch": 20.857217872980982, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.730926275253296, "target_z": -0.06591788812979126, "ce": 6.38797664642334, "genuineness": 4.413206577301025, "identity": 0.3207020163536072, "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_DRAM_reg/9f199732b39d0ba7f74fed87b5918673a2d951f07714fe1b0deeb9f621a0a6a4.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "9f4f94f6cabe813e0a70e44c786d01e8599e3c4e3dea3027df9497626958545f", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "a49733b18f5f3b8f4225c86c7ade83b78128746fb7f8505d24b64f06c20fb9c3", "generation_seconds_batch": 16.30342573300004, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.47939395904541, "target_z": 1.468749914986041, "ce": 5.589737892150879, "genuineness": 4.068024158477783, "identity": 0.760811448097229, "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_DRAM_reg/9f4f94f6cabe813e0a70e44c786d01e8599e3c4e3dea3027df9497626958545f.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "9f95518d08e92f3d262292fbd6028e646c2a8d72a5aa9c27a610bd231cc03571", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "9dd98e25aa709665cb3308006b238019e944f37b001e67ceffdfd77caaec19da", "generation_seconds_batch": 20.51925923593808, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.950666904449463, "target_z": 0.12695320224826084, "ce": 6.125868320465088, "genuineness": 4.494794845581055, "identity": 0.45359373092651367, "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_DRAM_reg/9f95518d08e92f3d262292fbd6028e646c2a8d72a5aa9c27a610bd231cc03571.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "9fb4df8ae62a20fa5baec60c3131273eadd08ff9664d536d2924b4361f536a18", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "c54d1e9f1933f6e6cb81ab4c7aae0bfbd64e0d20a1c8e9a626f49069dbe5234b", "generation_seconds_batch": 17.406035028980114, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.811868667602539, "target_z": 0.005126995683069635, "ce": 6.205295085906982, "genuineness": 4.582659721374512, "identity": 0.7179408073425293, "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_DRAM_reg/9fb4df8ae62a20fa5baec60c3131273eadd08ff9664d536d2924b4361f536a18.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": "9fb5718ee590234308ab75d26e9e821ad1b303aeaa305c8a3b48feb982b48b1f", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6c3ab8dbb16bf79351a97c993ee284eb57a3d00044edb3ca33b5a4329e71ff90", "generation_seconds_batch": 20.857154129073024, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.114457607269287, "target_z": 1.1484374068116219, "ce": 5.816103935241699, "genuineness": 3.6789097785949707, "identity": 0.525367021560669, "hypothesis": "De weg. Gaat linksaf naar de brug. Loop verder. Tot aan het kleine houten hek. Ik wacht aan de... Ik wacht aan de andere kant.", "wer": 0.23809523809523808, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_DRAM_reg/9fb5718ee590234308ab75d26e9e821ad1b303aeaa305c8a3b48feb982b48b1f.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": "9ffac0834903302dfe0c1d480c55a6e362758e29e5c78cbfb72e3d86023c1293", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The keys are on the kitchen table.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "70ed78c263fe82559a2ad47bb702871c355c6b8c78fe73facba77aa0d9e8a8ae", "generation_seconds_batch": 14.904511532979086, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.731719493865967, "target_z": 0.8124999084465057, "ce": 5.847874641418457, "genuineness": 4.18099308013916, "identity": 0.3869096636772156, "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_DRAM_reg/9ffac0834903302dfe0c1d480c55a6e362758e29e5c78cbfb72e3d86023c1293.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "a0190209c96588877733f953b1a57ee84ca727dce638f54f9ec8ac7c34fa204e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "2d41cf4d6bf5c83b811e6506c6c9060f0ef01391c3fcd915da02c45391220dae", "generation_seconds_batch": 21.255157538922504, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.535900115966797, "target_z": 0.6406250163488383, "ce": 5.5738525390625, "genuineness": 4.356721878051758, "identity": 0.372096985578537, "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_DRAM_reg/a0190209c96588877733f953b1a57ee84ca727dce638f54f9ec8ac7c34fa204e.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "a0e782011c82362b3d6cf18f517589d30c48da02692b938034f9698e39cf3cdb", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "7c5ac14c3f5dc7a3cdccf4f39c08b246e5cd94131f95ae931b7acfdc8541f088", "generation_seconds_batch": 20.732805949053727, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.586204528808594, "target_z": 1.5624998561302232, "ce": 5.871703147888184, "genuineness": 4.055472373962402, "identity": 0.5364933013916016, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude ramen. We kunnen daar samen in kijken.", "wer": 0.1111111111111111, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_DRAM_reg/a0e782011c82362b3d6cf18f517589d30c48da02692b938034f9698e39cf3cdb.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "a146b73711b7f367aa9fa43509af27c17f0e7a7018911d5ae0a58be08a00e661", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "2af615d818029dee66356ddaad4f1a969ce42ff868ff16777a10ababb3a33b1c", "generation_seconds_batch": 15.174126013065688, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6738638877868652, "target_z": 0.7617187949593053, "ce": 6.197352409362793, "genuineness": 4.331617832183838, "identity": 0.2891855537891388, "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_DRAM_reg/a146b73711b7f367aa9fa43509af27c17f0e7a7018911d5ae0a58be08a00e661.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "a1498a54b05bd2fc2bd35a367e97063b110ff1e38d394bd8bcd47d831f66e110", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "d0cbc7e2c513593fe010c2d3003722d1e308745ba6643cdefbfce9645e7489fb", "generation_seconds_batch": 21.75819694099482, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.455792188644409, "target_z": 0.5703125604907017, "ce": 5.7128496170043945, "genuineness": 4.557555198669434, "identity": 0.8518565893173218, "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_DRAM_reg/a1498a54b05bd2fc2bd35a367e97063b110ff1e38d394bd8bcd47d831f66e110.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "a24f615d3f5a24df485b03ab59c35b20dd9062fc3e47390cf9ce4fb810761ca0", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "a2ba952e361a7723dc4dbd582208e566bf4d3cbd5e873620c9aa12bf92e18c60", "generation_seconds_batch": 14.972811780055054, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.286675214767456, "target_z": 0.42187494441394985, "ce": 6.292664527893066, "genuineness": 4.482243061065674, "identity": 0.7541885375976562, "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": 0.03999999999999915, "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_DRAM_reg/a24f615d3f5a24df485b03ab59c35b20dd9062fc3e47390cf9ce4fb810761ca0.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "a2c3ac01efc1c9368f8a6b549bd8d14ddb2a719a6d63c6c56b4339da101ad0a2", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "d45746277ec04cc5a876000b1b13afe8a0bdd18b358089419f7ae79a1a59873c", "generation_seconds_batch": 20.789070877013728, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.119783639907837, "target_z": 0.27539060905988266, "ce": 6.467403411865234, "genuineness": 4.557555198669434, "identity": 0.45353764295578003, "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_DRAM_reg/a2c3ac01efc1c9368f8a6b549bd8d14ddb2a719a6d63c6c56b4339da101ad0a2.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "a3a0e7701900e401deefef02387a6a012249fb86b9030e262139570e7ea21e78", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "cc8401cc5a3e9f195e37d12bbd4440927ba21653854e4a091fd52b85a16108a8", "generation_seconds_batch": 15.147895290981978, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7183682918548584, "target_z": 0.8007812704360479, "ce": 5.649308204650879, "genuineness": 4.406930446624756, "identity": 0.7803313732147217, "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_DRAM_reg/a3a0e7701900e401deefef02387a6a012249fb86b9030e262139570e7ea21e78.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "a48b74f8bff74fa63590d3061b2b92854e168f6bab58ad267021bc59123ff924", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "b815882679898d4b87da6ffea30f153e7def05f0b44236ce03e1079743fda7a2", "generation_seconds_batch": 20.87778265099041, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.998746156692505, "target_z": 1.046874970572091, "ce": 6.1894097328186035, "genuineness": 4.042920112609863, "identity": 0.2478957176208496, "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_DRAM_reg/a48b74f8bff74fa63590d3061b2b92854e168f6bab58ad267021bc59123ff924.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "a48ee924cdeb4e55d66f963ad15716b6169235694601b4b13151febe594b798d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "e27febd54c90ab99379456c97a0863f59a273195ace3423d66a4834a5688865a", "generation_seconds_batch": 17.406035028980114, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.291125774383545, "target_z": 0.42578129659418906, "ce": 6.109982967376709, "genuineness": 4.287685871124268, "identity": 0.5831618905067444, "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_DRAM_reg/a48ee924cdeb4e55d66f963ad15716b6169235694601b4b13151febe594b798d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "a5c859771b357f78a24b19ae4d46a827af0b4c850fe7c1677208acc5a65babbe", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "217a6b8bea417090d81b54d9333151d8546453f0f937be6f78e7492ba54a540c", "generation_seconds_batch": 20.879176541930065, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.764222145080566, "target_z": 1.7187497580371935, "ce": 5.530168056488037, "genuineness": 3.5345609188079834, "identity": 0.22050052881240845, "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_DRAM_reg/a5c859771b357f78a24b19ae4d46a827af0b4c850fe7c1677208acc5a65babbe.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "a614b6254fcfce9592dd088cff037272ae481613d2c4218aa4a7416935b81a37", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "2aabeb5f1efe899c8272558947b541ff95c3a0716a4ee160ee28f4b532326bcf", "generation_seconds_batch": 16.24833213200327, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.721468925476074, "target_z": -0.07421880803837588, "ce": 6.038498878479004, "genuineness": 4.65797233581543, "identity": 0.7794536352157593, "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_DRAM_reg/a614b6254fcfce9592dd088cff037272ae481613d2c4218aa4a7416935b81a37.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "a6a3609edb9f8984aef802bad89af8fbc8dd201e56763b511fa8d50d46444ec7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "1b0903f4f1ced75c1824bdf72fc3b3b06385b881d079ec354d27d7794f2bb5f1", "generation_seconds_batch": 22.279276649001986, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.408186912536621, "target_z": 1.406249954223253, "ce": 6.165581703186035, "genuineness": 4.1998209953308105, "identity": 0.534618616104126, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude, het oude raam. We kunnen er samen in kijken.", "wer": 0.1111111111111111, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_DRAM_reg/a6a3609edb9f8984aef802bad89af8fbc8dd201e56763b511fa8d50d46444ec7.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "a6e3b4fab1156560da5f782526f8d4ac1205f6f333f4f36ece3551e6290c5ab8", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "a32f5e087075f328a6d9762f238726a9d17f491d33817f782b640be477f5a907", "generation_seconds_batch": 22.756419694051147, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.678314208984375, "target_z": 0.7656249378744145, "ce": 5.847874641418457, "genuineness": 4.494794845581055, "identity": 0.34080770611763, "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_DRAM_reg/a6e3b4fab1156560da5f782526f8d4ac1205f6f333f4f36ece3551e6290c5ab8.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": "a6f8f209dbe41f01cced8b455f9ba0212ee9ade1190ec92c4b9a5e001513baa4", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "98f4bd8023dd3707c475d5785128f831f3852876a247ab19a4bc9dc840c2e51c", "generation_seconds_batch": 15.377575946971774, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6649630069732666, "target_z": 0.7539062998639567, "ce": 5.363371849060059, "genuineness": 4.256305694580078, "identity": 0.24042782187461853, "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_DRAM_reg/a6f8f209dbe41f01cced8b455f9ba0212ee9ade1190ec92c4b9a5e001513baa4.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "a784d30eeddf86d06dad6942bf11f7e4c943b359a0c692fd10270b5e831248b1", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "e9332f4e6585cc77816ce608d99bb6b6eca2f3c43714e5a3a9be83e2f45a1dae", "generation_seconds_batch": 16.128198569989763, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0597026348114014, "target_z": 0.2226562148499977, "ce": 5.871703147888184, "genuineness": 4.319066047668457, "identity": 0.5869458913803101, "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_DRAM_reg/a784d30eeddf86d06dad6942bf11f7e4c943b359a0c692fd10270b5e831248b1.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "a79f2424c075e3c2eaa7bdbda9861ea2dcc8119c02fb9b760a2e0cf48b12b182", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "a1c1631fe36e5d38241ef0849a6190bfc8ce21ed38530a7acb1d81ce0fcb10fd", "generation_seconds_batch": 21.232470880029723, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3934860229492188, "target_z": 0.515625094823262, "ce": 6.181467056274414, "genuineness": 4.319066047668457, "identity": 0.5310988426208496, "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_DRAM_reg/a79f2424c075e3c2eaa7bdbda9861ea2dcc8119c02fb9b760a2e0cf48b12b182.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "a818a9a425965d2ddf7a0d8de7eb3b5b5bb872d12b11d01d94b43aad61fd604c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "37b04b9009f8164c403cb90963a9a18d873a0f1c248c5798bf891a8ac081c860", "generation_seconds_batch": 20.729799936059862, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.150061130523682, "target_z": 1.1796873871930158, "ce": 5.232317924499512, "genuineness": 4.20609712600708, "identity": 0.24263666570186615, "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_DRAM_reg/a818a9a425965d2ddf7a0d8de7eb3b5b5bb872d12b11d01d94b43aad61fd604c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "a88a4829785ad94ee707c998fc75cdd9e22618d191f9b85f1e5b56c62962b36e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "8ab2940c613e4a342dfb428b74ce05a3e131a30e1a08e43ebbfdeba57585cc4e", "generation_seconds_batch": 22.21201195497997, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.550601005554199, "target_z": 1.531249875748829, "ce": 5.121120452880859, "genuineness": 3.7793264389038086, "identity": 0.4531693458557129, "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_DRAM_reg/a88a4829785ad94ee707c998fc75cdd9e22618d191f9b85f1e5b56c62962b36e.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "a88af0e6289035b709c3bcc444e3e01046ba37c2ddb1428ce3217d3c255766eb", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "dc348f48e3fbecef7af2a0b39df1741aa3c147feebe531b66362b9f663a167fd", "generation_seconds_batch": 21.53801813791506, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.139810562133789, "target_z": 0.29296867070813437, "ce": 6.348263263702393, "genuineness": 4.394378662109375, "identity": 0.42093196511268616, "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_DRAM_reg/a88af0e6289035b709c3bcc444e3e01046ba37c2ddb1428ce3217d3c255766eb.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "a8b357c599b7def3ca1a4359743809a059fe267fbd3fafdd4a413dd02fa34d62", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "3f30d655a30ee62dcc4294bc9801ba4455ccbefa39c9c81ecdd8d268d7c95b3b", "generation_seconds_batch": 20.518998953979462, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.532799243927002, "target_z": 1.515624885558132, "ce": 5.649308204650879, "genuineness": 3.9550557136535645, "identity": 0.49156099557876587, "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_DRAM_reg/a8b357c599b7def3ca1a4359743809a059fe267fbd3fafdd4a413dd02fa34d62.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "a68135bedec820f4c04013aab5a25e832965b5ba3c62b57397dd790f3d15964d", "generation_seconds_batch": 16.249565552105196, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.542101502418518, "target_z": -1.1093750359674441, "ce": 6.205295085906982, "genuineness": 4.18726921081543, "identity": 0.6223819255828857, "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_DRAM_reg/a8bd001b1fcafd2f49b65d73e6ed12567e4c9664909a4d08700c9febfd977540.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "a9147ec187ba940652087426ff1efcbd5c4555516a99dc771e4e9442a3d9c3aa", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "4db4e99ecf390f9b08a7e115c31482d8860f2781bf3a6df6f80a6734067008e2", "generation_seconds_batch": 21.592677181004547, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.425988674163818, "target_z": 1.4218749444139498, "ce": 5.2065043449401855, "genuineness": 4.005263805389404, "identity": 0.2499646693468094, "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_DRAM_reg/a9147ec187ba940652087426ff1efcbd5c4555516a99dc771e4e9442a3d9c3aa.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "f356599dabdd33069bf1e3cb1a86429c76564d9cea22cdd868094c5032df4334", "generation_seconds_batch": 14.92464466392994, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.81353759765625, "target_z": 0.0065918515925181016, "ce": 6.221180438995361, "genuineness": 4.287685871124268, "identity": 0.5757216215133667, "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_DRAM_reg/a94b7d5da78ff217f8e5a4f7ae4520900bef0ad764d0e4ee898e69de10c28aea.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "aa29e1969651b8d5a3fee701a5202073c7210510b6fb49737a1ce98dbaca38cd", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "51a953bb0ea50c2c80617636cb84f9438e8cd5c0200f80e50ecc456756ab47a1", "generation_seconds_batch": 21.352475182968192, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3445310592651367, "target_z": 0.4726562671662802, "ce": 5.9431867599487305, "genuineness": 4.243753433227539, "identity": 0.4989578127861023, "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_DRAM_reg/aa29e1969651b8d5a3fee701a5202073c7210510b6fb49737a1ce98dbaca38cd.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": "aa33e637c4b5d77cf52daf12b0d171c08b1480fdf6a36de9d07cf8943bbc4da4", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "c66f6e26375eb77a9e6e211c21539b614833efb9a69f712bd0d642e1e6a9a656", "generation_seconds_batch": 15.209956906037405, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.089743137359619, "target_z": 0.24902341195494018, "ce": 6.022613525390625, "genuineness": 4.419482707977295, "identity": 0.4418324828147888, "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_DRAM_reg/aa33e637c4b5d77cf52daf12b0d171c08b1480fdf6a36de9d07cf8943bbc4da4.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "aaa8bf200f56ba9bfd4684fd2e257f6e8a70e11b0bebc3dab28ec3ebff663ae2", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "6627fbdafb2c4ab8920e76a77c7902f24bf643f8ad066acece05f89c1f6f56c2", "generation_seconds_batch": 16.00596959097311, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6675074100494385, "target_z": -0.12158199415857317, "ce": 6.340320587158203, "genuineness": 4.570107460021973, "identity": 0.5692428350448608, "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_DRAM_reg/aaa8bf200f56ba9bfd4684fd2e257f6e8a70e11b0bebc3dab28ec3ebff663ae2.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "aafd89c02d61c4a523e74081bcf0bf807fec101cd9336a3ac5e8101f777a6443", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "802a000b750ae1e875de671601345bb56e680d0d84d48cb4cb5cfb80e0ed20f0", "generation_seconds_batch": 20.45422521501314, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.069953441619873, "target_z": 1.1093751406000092, "ce": 5.824047088623047, "genuineness": 4.432034492492676, "identity": 0.753380537033081, "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_DRAM_reg/aafd89c02d61c4a523e74081bcf0bf807fec101cd9336a3ac5e8101f777a6443.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "ab4bc1c49af9cdd710660234c5c57a94607f6fe62c6b8d07a935424045db5fd2", "bucket": "vn_S_DRAM_reg", "phase": "wording", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "bb02ad4c846794001164c841b985325c1ccd57699cdafef5314280a6c14ee615", "generation_seconds_batch": 16.127483104006387, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4944963455200195, "target_z": -0.27343753760232803, "ce": 6.1894097328186035, "genuineness": 4.770940780639648, "identity": 0.5404447317123413, "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_DRAM_reg/ab4bc1c49af9cdd710660234c5c57a94607f6fe62c6b8d07a935424045db5fd2.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "ab733fc74e0f8baae7cd556a6c705c5e856c7a3d94eda812c625727a95340735", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "7caf149f9a945c6446cfae470dea5447f58a82badffcf50a748cdd777e9663d3", "generation_seconds_batch": 15.901571974041872, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9542417526245117, "target_z": 1.0078124950953484, "ce": 5.442798614501953, "genuineness": 4.287685871124268, "identity": 0.4317099452018738, "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_DRAM_reg/ab733fc74e0f8baae7cd556a6c705c5e856c7a3d94eda812c625727a95340735.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "5ceaeb98f4caf7e6378f8bef687b11a08e2e48692f714ce5beab7a0a7f39fbb1", "generation_seconds_batch": 14.661946450010873, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.60241961479187, "target_z": -0.17871095609680354, "ce": 6.41180419921875, "genuineness": 4.419482707977295, "identity": 0.45401471853256226, "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_DRAM_reg/ab7a94299f968552a7b966fa3391724b652d8e20483888a7767196f29a7f4a69.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "abe739d32dec9b1bf35f5335feb2a2052f8d86ae147257832458ead2d7021288", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "ee715b583d9672cc819c61c823548d904fc73a0c31a4a805e36ee5cda18d47c1", "generation_seconds_batch": 20.994715513894334, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.09665584564209, "target_z": 1.1328124166209248, "ce": 5.657250881195068, "genuineness": 4.080576419830322, "identity": 0.5770388245582581, "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_DRAM_reg/abe739d32dec9b1bf35f5335feb2a2052f8d86ae147257832458ead2d7021288.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "ac896a08c1d19dcb9026406dc4e62e297b1d667c61cf461927bfbdf4d44864b7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "d6083b3fbef4ed466bc64ddd8074996ff4d1aec4cf4491252a3663b9782078be", "generation_seconds_batch": 21.440973090939224, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.848306655883789, "target_z": 0.03710940401918794, "ce": 6.403861999511719, "genuineness": 4.306513786315918, "identity": 0.5905627608299255, "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_DRAM_reg/ac896a08c1d19dcb9026406dc4e62e297b1d667c61cf461927bfbdf4d44864b7.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "ac8a73f5844c7f1a64e9386557a18c109e635e2f42816612d99c849c83d06654", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "623b364e364df1b4c4ad8cacb42f7cbcb0fd7423141f884019d4e81ddcfdf4e0", "generation_seconds_batch": 21.87371906498447, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0018467903137207, "target_z": 0.17187489209766738, "ce": 6.268836498260498, "genuineness": 4.545003414154053, "identity": 0.5826672315597534, "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_DRAM_reg/ac8a73f5844c7f1a64e9386557a18c109e635e2f42816612d99c849c83d06654.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "acfbf75793c647f48ef49e93049559e64cc35c308a7aeb51f84a3c9b36d4dc94", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "7e3d913cbff606d52507ac42042699da9ecf055092c242e81c8fdcf732eaf5e2", "generation_seconds_batch": 21.54336655093357, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0129730701446533, "target_z": 0.18164066791570047, "ce": 6.221180438995361, "genuineness": 4.582659721374512, "identity": 0.3263320326805115, "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_DRAM_reg/acfbf75793c647f48ef49e93049559e64cc35c308a7aeb51f84a3c9b36d4dc94.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "ad045039a5f0bdba835ce03036b81ec62d8f56f9b258d7026dd81bb95491f902", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "34c692aa7adc230295c7a43b7148fbbfe528513a0f73570696d0bb9339635f54", "generation_seconds_batch": 22.35787030204665, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.827723264694214, "target_z": 0.01904291755770015, "ce": 6.213237762451172, "genuineness": 4.2374773025512695, "identity": 0.43307268619537354, "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_DRAM_reg/ad045039a5f0bdba835ce03036b81ec62d8f56f9b258d7026dd81bb95491f902.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "ad7a9d1f9bc9d66f10c6ae03671d4a12b5a1be4e87fea95a04c91ec50ef5ef2a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "a435d078ea2a7aae0ada15483997cb74916f5a9669523dd8f9bf8058c9b08ec2", "generation_seconds_batch": 20.994715513894334, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8340797424316406, "target_z": 0.9023437066755786, "ce": 5.577824115753174, "genuineness": 4.18099308013916, "identity": 0.5591917634010315, "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": 0.03999999999999915, "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_DRAM_reg/ad7a9d1f9bc9d66f10c6ae03671d4a12b5a1be4e87fea95a04c91ec50ef5ef2a.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "bec71cf34bfe4f55330251008682d68b09caed0f36c08a6e70249eb3d3be05cd", "generation_seconds_batch": 16.21426491998136, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9515423774719238, "target_z": -0.7499999476837175, "ce": 6.078212261199951, "genuineness": 4.494794845581055, "identity": 0.6112931370735168, "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_DRAM_reg/adcc0121d6c11de6953f8243f84ad890832f75283c96342c1db246fddc443438.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "ebd39324fe809a372ee4705f04e89d6f636ca6c5528dd5f2ed4d3bdbb4e9441a", "generation_seconds_batch": 13.414194418932311, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2755491733551025, "target_z": 0.4121093778610467, "ce": 6.340320587158203, "genuineness": 3.948779582977295, "identity": 0.45193254947662354, "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_DRAM_reg/ae3130a5db36c2790a31e23135fde257111c56fa9c5b8061c1f3e51a8cb61258.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "aebca20d0e2ac2db48588a4a3ca855f3ba00f1a9701f7394b53d1b7f61b0de45", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "40316dab229bcc2a33f4418a4ec06716be444869062dd9ff6298acb5c1d65186", "generation_seconds_batch": 22.337924418970942, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6625006198883057, "target_z": -0.12597656188691855, "ce": 6.23706579208374, "genuineness": 4.319066047668457, "identity": 0.6091057062149048, "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_DRAM_reg/aebca20d0e2ac2db48588a4a3ca855f3ba00f1a9701f7394b53d1b7f61b0de45.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": "aec5aa0f9c15500290948e753a57be0f5e406da249d6de29d8d48125d46b6184", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Someone moved the little blue chair.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "61fbcc7eb3c2d87118ab18804d3c40e38cdf6bc52e28aaf65fc322fdd8c7b043", "generation_seconds_batch": 15.313343288027681, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9898452758789062, "target_z": 1.0390624754767426, "ce": 5.9431867599487305, "genuineness": 3.973883628845215, "identity": 0.1990608274936676, "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_DRAM_reg/aec5aa0f9c15500290948e753a57be0f5e406da249d6de29d8d48125d46b6184.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "aecfc7165a42516f5751e57959be1a8690e120c0bdc4c0dd2e51c7d400709cd1", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "4bcea4e2add05b5b46d8485bc7c142cff3ae1fa4c35fc2c4ef639510a4f41a29", "generation_seconds_batch": 20.6980390480021, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.651374578475952, "target_z": -0.13574212843982172, "ce": 6.2767791748046875, "genuineness": 4.118232727050781, "identity": 0.44360169768333435, "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_DRAM_reg/aecfc7165a42516f5751e57959be1a8690e120c0bdc4c0dd2e51c7d400709cd1.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": "af6b0ee3e60540898cc3247976977ddee265407e2d8cd6e1f2601cc254430b2c", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The keys are on the kitchen table.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "70bb968fda5f945f0b147055c4fe607b31c9e3ed16630256a0dc573da73bd5f1", "generation_seconds_batch": 22.325334085035138, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8563320636749268, "target_z": 0.9218750490465148, "ce": 6.070269584655762, "genuineness": 4.18099308013916, "identity": 0.28937360644340515, "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_DRAM_reg/af6b0ee3e60540898cc3247976977ddee265407e2d8cd6e1f2601cc254430b2c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "af8e792c8783ee43ab08c1647a3b31efd4412fcc12692e708bcd5ac7c6199e7f", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "891dbe963e8b33cf7d2032f62c6019c0bcd96b8a34f6f878d04cdb3aa058d936", "generation_seconds_batch": 21.96490627503954, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.58040452003479, "target_z": 0.6796874918255809, "ce": 6.109982967376709, "genuineness": 4.482243061065674, "identity": 0.3992360830307007, "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_DRAM_reg/af8e792c8783ee43ab08c1647a3b31efd4412fcc12692e708bcd5ac7c6199e7f.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "afb24dd66da35da8876e089a3eaf57262a1d17a13c2f98a3f89c8d066d0f82b3", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "b5cdd2cf5c36dbed0c9130487ba9a7e0fb7a934ded1c7c9398fe9fbf32724738", "generation_seconds_batch": 21.150260988040827, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6686198711395264, "target_z": -0.1206055630623608, "ce": 5.80021858215332, "genuineness": 4.444586753845215, "identity": 0.6654809713363647, "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_DRAM_reg/afb24dd66da35da8876e089a3eaf57262a1d17a13c2f98a3f89c8d066d0f82b3.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "b0887580b5b6f55e600ad842b97337499e0354fbca2b0515b98bec00e2b8269e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "2e16bfa9c6e0fb4c30904e5ca9f837533fc5fb6c4b208e3d9786d300bd61825c", "generation_seconds_batch": 15.734572071931325, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8138155937194824, "target_z": 0.006835854734006232, "ce": 6.062326908111572, "genuineness": 4.632867813110352, "identity": 0.619249701499939, "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_DRAM_reg/b0887580b5b6f55e600ad842b97337499e0354fbca2b0515b98bec00e2b8269e.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "b0ae505f092570a6d9669aa1486f310771f3ccfc4616930357270adca8096b96", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "8adc5261d457ad61765bf1d0f3f5fa7b47d08a988dce32426f1ea853af616fef", "generation_seconds_batch": 15.227305427892134, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.411287784576416, "target_z": 0.531250085013959, "ce": 5.847874641418457, "genuineness": 4.6203155517578125, "identity": 0.712940514087677, "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_DRAM_reg/b0ae505f092570a6d9669aa1486f310771f3ccfc4616930357270adca8096b96.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "b0ffb27153d9193eef0db1f1fabab86395ef530b0958517c562168980dfafd93", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "d074eb496592a4d0381b0bb8a837ee6120a9ebe5f5aaa3e8750973030035de79", "generation_seconds_batch": 15.200643323012628, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0875179767608643, "target_z": 0.24707034049738555, "ce": 6.316492557525635, "genuineness": 4.7332844734191895, "identity": 0.5343176126480103, "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_DRAM_reg/b0ffb27153d9193eef0db1f1fabab86395ef530b0958517c562168980dfafd93.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "b11260ab1d6fb20adefe5ce79cbf3d55c765a957430fa140da4305f8a96a5b65", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "3b77b736cbc3579f5467ea041d64bdaaf8762208c5bc92a2cc1fc5b2481a4ad9", "generation_seconds_batch": 20.410747074056417, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.869683265686035, "target_z": 0.9335936870569727, "ce": 6.125868320465088, "genuineness": 4.6705241203308105, "identity": 0.6318544149398804, "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_DRAM_reg/b11260ab1d6fb20adefe5ce79cbf3d55c765a957430fa140da4305f8a96a5b65.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "b11acc2386ba5e49487ea4d53311d1303ab2fced2a30377fdb3dd68c9366dbc7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "4e72232815aa6ff6f855f282bf5c9de1a4b1c16696f5d3df63bb45c7fd915493", "generation_seconds_batch": 21.29505303595215, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5657033920288086, "target_z": -0.2109375768395399, "ce": 6.340320587158203, "genuineness": 4.369274139404297, "identity": 0.45775851607322693, "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_DRAM_reg/b11acc2386ba5e49487ea4d53311d1303ab2fced2a30377fdb3dd68c9366dbc7.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": "b1b1fc048d95e6399bec5e60947b0cc956fba02571c1bd966231de5da036caa8", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "3b78724198813ea080e60555654b15b4e7902b9c574f00008dda59c3963c6795", "generation_seconds_batch": 15.714376059942879, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.141160488128662, "target_z": 1.1718751013627973, "ce": 5.601652145385742, "genuineness": 4.174716949462891, "identity": 0.7133852243423462, "hypothesis": "Quelqu'un a déplacé la petite chasse bleue. Elle était ici, près de la grande bibliothèque. Nous la remettrons après le déjeuner.", "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_DRAM_reg/b1b1fc048d95e6399bec5e60947b0cc956fba02571c1bd966231de5da036caa8.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "b1bb8e0d24d292e20332788a3e0752a90798838d7fb60a1b81532738e6397783", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "0e1a8ee7a05108371d02938a9a3925633d62e2ed52cded005bc19b2be9eee5be", "generation_seconds_batch": 16.03295825200621, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.048576593399048, "target_z": 0.21289064829709453, "ce": 5.967014789581299, "genuineness": 4.400654315948486, "identity": 0.8135873675346375, "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_DRAM_reg/b1bb8e0d24d292e20332788a3e0752a90798838d7fb60a1b81532738e6397783.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "b210d0f17ccb08bf13e51523b8b1f97619b45160461b7c9ffa9bb3032a0af191", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "b182b84b7f47e6f5915828fd3e82955ea9ad6cf9eae49711c8a4b3bf56782889", "generation_seconds_batch": 20.26349670998752, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.94566011428833, "target_z": 0.12255863451991544, "ce": 6.117925643920898, "genuineness": 4.344170093536377, "identity": 0.5576878786087036, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/b210d0f17ccb08bf13e51523b8b1f97619b45160461b7c9ffa9bb3032a0af191.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "b2801d75452802fc5242280bbd5b3804e21fb4110fbb522ed6b050d905876f5f", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "d576c0dae409665bac03a163c19282209b9511c29ab26bf093c6b2c4fa8dcac6", "generation_seconds_batch": 21.071550339926034, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.194565773010254, "target_z": 1.2187500719348885, "ce": 5.744620323181152, "genuineness": 4.243753433227539, "identity": 0.2910397946834564, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/b2801d75452802fc5242280bbd5b3804e21fb4110fbb522ed6b050d905876f5f.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "b29a1a40f01a5f3beae807c26eb3cfb8e1e4cbf4ac23033ebd42036bd18d46f9", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "356a9ec205fb315511fc434b003caef947033afb3dc612c0a858121a1dc44467", "generation_seconds_batch": 14.919987099943683, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3289544582366943, "target_z": 0.4589843484331378, "ce": 6.332377910614014, "genuineness": 4.218649387359619, "identity": 0.2180958092212677, "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_DRAM_reg/b29a1a40f01a5f3beae807c26eb3cfb8e1e4cbf4ac23033ebd42036bd18d46f9.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "b309f414864565588d4e806330fe7887066e7c242eb72cc0ef63fd9eff5fa024", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "92eb6c3285c454629d66079a8891edec254720aa62b9eebe16fb7c5c5e0215b1", "generation_seconds_batch": 19.914986506919377, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7139177322387695, "target_z": 0.7968749182558086, "ce": 5.704906940460205, "genuineness": 4.582659721374512, "identity": 0.5021477937698364, "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_DRAM_reg/b309f414864565588d4e806330fe7887066e7c242eb72cc0ef63fd9eff5fa024.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "b3bbab0afe3e9cf0d842eade483777d3bc17cb67b3c47871fb07c6f56196483f", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "8468a9044747f23b1e298f6677937aa21e0a1a5253a7f3921b610f3318e944b9", "generation_seconds_batch": 16.51156911405269, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.900599241256714, "target_z": 0.08300773422993676, "ce": 6.570657730102539, "genuineness": 4.65797233581543, "identity": 0.31696414947509766, "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_DRAM_reg/b3bbab0afe3e9cf0d842eade483777d3bc17cb67b3c47871fb07c6f56196483f.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "b3fabae80f9717ba12942cbaee38697990768245b0a8ee7004505a62c1624fa7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "3c2ef2692e4827cd45ce381127b5d044f91aac366f9adfae5e1c3668fa26e3b3", "generation_seconds_batch": 19.886398899019696, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.268873453140259, "target_z": 0.40624995422325283, "ce": 6.340320587158203, "genuineness": 4.306513786315918, "identity": 0.43395984172821045, "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_DRAM_reg/b3fabae80f9717ba12942cbaee38697990768245b0a8ee7004505a62c1624fa7.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "cbb3f2b9830e94802654c4540ddf6a9710c070f3d2ebf3e9ec1754f7711df9a6", "generation_seconds_batch": 15.267393877031282, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.412163257598877, "target_z": -0.34570306491801933, "ce": 6.332377910614014, "genuineness": 4.344170093536377, "identity": 0.5363434553146362, "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_DRAM_reg/b429ad869b4f073af736e4e6510807542fc4f5324b0eb7f096c719d9002f4f71.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": "b438992655495ab77235a2f334ab8784140200c061155e24d8db4207e4e6e9ee", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "aefcfcb1a345c1d6efc82437a966c890e16b342fc027cd550bca5792b9ce4d57", "generation_seconds_batch": 20.403864764026366, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.847431182861328, "target_z": 0.9140625539511663, "ce": 5.414999485015869, "genuineness": 4.18726921081543, "identity": 0.3908035159111023, "hypothesis": "Die Straße führte hinter der Brücke nach links. Je weiter bis zu dem 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_DRAM_reg/b438992655495ab77235a2f334ab8784140200c061155e24d8db4207e4e6e9ee.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "b51d2d17b014b2014254269bc3c558eea0e94819a5e0c5f5c966bd084c642605", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "23c3a3c7d4a242622375ff05dd918db0ac22f4aae238f8d87f027767dc09ef6d", "generation_seconds_batch": 21.76063445804175, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7851247787475586, "target_z": 0.8593748790185968, "ce": 6.149696350097656, "genuineness": 4.20609712600708, "identity": 0.16918499767780304, "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_DRAM_reg/b51d2d17b014b2014254269bc3c558eea0e94819a5e0c5f5c966bd084c642605.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": "b564b52c02b60a436c871abf63892ac3f1387526557a298686b966c1ebc7d356", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Escribí el número en el sobre.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f1495742558683b090703f1e35662032234713a0eca3f0b9e40767abd42144a0", "generation_seconds_batch": 23.087157826987095, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.08775520324707, "target_z": 1.1250001307907063, "ce": 5.39911413192749, "genuineness": 4.030367851257324, "identity": 0.3518078327178955, "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_DRAM_reg/b564b52c02b60a436c871abf63892ac3f1387526557a298686b966c1ebc7d356.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "b5a77c4971f0220881634cb708d5dabde97d947c9c3b46ae9950e4bcc9a2b0b3", "bucket": "vn_S_DRAM_reg", "phase": "wording", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "ee94aa9e087c77fddd17e7d8905961d668146358be634bc346bd9dfe38824bdb", "generation_seconds_batch": 14.779266525059938, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0271997451782227, "target_z": -0.6835938440058201, "ce": 6.38797664642334, "genuineness": 4.6203155517578125, "identity": 0.496345579624176, "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_DRAM_reg/b5a77c4971f0220881634cb708d5dabde97d947c9c3b46ae9950e4bcc9a2b0b3.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "b5e5c1bd39f52eb4232f93eddaab8b5676fdc09d1f953f6d19c1545ae8dbb3d4", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "f1628606de3e8cc4ea134c7f376169157f99b4c14a07726c8d29c83190b1479b", "generation_seconds_batch": 21.352475182968192, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8922548294067383, "target_z": 0.07568366394782435, "ce": 6.070269584655762, "genuineness": 4.099404335021973, "identity": 0.40199801325798035, "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_DRAM_reg/b5e5c1bd39f52eb4232f93eddaab8b5676fdc09d1f953f6d19c1545ae8dbb3d4.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "b5f48268c706d6d704f969d798f2c181373c1163b0ca96507aac1eab93afc582", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "019161b584a1e96c2d4c50d41a076a89035626d6d9e95ba2204bddf4c0816942", "generation_seconds_batch": 21.2666702340357, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3489813804626465, "target_z": 0.4765624100813895, "ce": 6.062326908111572, "genuineness": 4.406930446624756, "identity": 0.49996766448020935, "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_DRAM_reg/b5f48268c706d6d704f969d798f2c181373c1163b0ca96507aac1eab93afc582.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "b5f699fb2dc11e01b9efc1eb9705f0d85c2631041ae3b26b94633f413edc6bf6", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "a9534aa9a5c1ffc7f9c0a9fb023c2b0b3f97f005821557d07d9ec6e63cb92c57", "generation_seconds_batch": 21.018774896045215, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5211989879608154, "target_z": -0.2500000523162825, "ce": 5.95112943649292, "genuineness": 4.281409740447998, "identity": 0.7784230709075928, "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_DRAM_reg/b5f699fb2dc11e01b9efc1eb9705f0d85c2631041ae3b26b94633f413edc6bf6.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "b6c822b51b14c919dbd4eee1c0769ecf9fb1586858991e6889112067468507b4", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "b34cae170a3e40df577c22e1927ea17fbc8e64e4a1342c6c5cf85fa62e99e8e5", "generation_seconds_batch": 20.377851914032362, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0096352100372314, "target_z": 0.17871095609680354, "ce": 5.990842819213867, "genuineness": 4.582659721374512, "identity": 0.7276744842529297, "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_DRAM_reg/b6c822b51b14c919dbd4eee1c0769ecf9fb1586858991e6889112067468507b4.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "b6d5b21ebe2697662ed06da5626b0b3eef807a5afe242d7fa3082fc9aab2cddb", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "c1d431781a008168c5b8ed36cd6734e565e33007f09e7c056af446ebe9f2c7be", "generation_seconds_batch": 19.26112615596503, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9328651428222656, "target_z": 0.11132821205756381, "ce": 6.23706579208374, "genuineness": 4.720732688903809, "identity": 0.5007306933403015, "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_DRAM_reg/b6d5b21ebe2697662ed06da5626b0b3eef807a5afe242d7fa3082fc9aab2cddb.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "b783348eb501b959c1a50678c55f582377554bfdc5a17ed3b3ec357352280901", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "4f5f3c8c77c897c6d74fda9adeab1d42264424765d2a62e6153c93249eb58865", "generation_seconds_batch": 22.145443430985324, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.141160488128662, "target_z": 1.1718751013627973, "ce": 6.0146708488464355, "genuineness": 4.168440818786621, "identity": 0.5235830545425415, "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_DRAM_reg/b783348eb501b959c1a50678c55f582377554bfdc5a17ed3b3ec357352280901.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": "b82d0e4a3c125cc378f443b3740ed9b66f96e1099a79ed334d3de2061a5677df", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "b11172ddae7ffdd99332fdbb50c6483c78622503d70f64eb62d03bfb96eca7e2", "generation_seconds_batch": 15.383068852010183, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.820728302001953, "target_z": 0.8906248593999908, "ce": 5.756534099578857, "genuineness": 4.444586753845215, "identity": 0.6804615259170532, "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_DRAM_reg/b82d0e4a3c125cc378f443b3740ed9b66f96e1099a79ed334d3de2061a5677df.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "b91e53f7321683beec0e4a993175b8f4d8a87587d3ec58af1bbaa61eadc31567", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "ecb52fa4c706db7926dc942cb5d8ba05a8e70b13843bb83132fc041d45f8e38b", "generation_seconds_batch": 21.890862816013396, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6694133281707764, "target_z": 0.757812442779066, "ce": 6.038498878479004, "genuineness": 3.9676074981689453, "identity": 0.2571766972541809, "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_DRAM_reg/b91e53f7321683beec0e4a993175b8f4d8a87587d3ec58af1bbaa61eadc31567.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "b92a2af5096fab1f451e793da18d46c141a34812ae3c433bd77c5ca68a65d18a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "5e7e218a75713ec1d6922fbfb9a764c303754f3ae8fc63d7c68a3bbc557c8341", "generation_seconds_batch": 16.19155024900101, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9339776039123535, "target_z": 0.11230464315377618, "ce": 5.967014789581299, "genuineness": 4.821148872375488, "identity": 0.4145504832267761, "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_DRAM_reg/b92a2af5096fab1f451e793da18d46c141a34812ae3c433bd77c5ca68a65d18a.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "c22c6dd7dc1c6690210eabe98c7e3dc68d6dffa29ecd98f278d0d1bb10504290", "generation_seconds_batch": 15.315051507088356, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.363208293914795, "target_z": -0.3886718925750011, "ce": 6.1894097328186035, "genuineness": 4.419482707977295, "identity": 0.6450133919715881, "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_DRAM_reg/ba70e7e208a6b15a9236ccd5a69e0e5fd403df2a937202d690effa2cdc45d9f2.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "bac0d8e87ab5e9e79f623846bb36571c4efa02ed6dc13fc307f2d87c02cf3327", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "f9e6f7e73b2de468e4502cd9ee283e25b3a1678bb35ee28bbcaef963ed00f87e", "generation_seconds_batch": 16.03295825200621, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0585899353027344, "target_z": 0.2216795744886554, "ce": 6.1020402908325195, "genuineness": 4.406930446624756, "identity": 0.45256567001342773, "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_DRAM_reg/bac0d8e87ab5e9e79f623846bb36571c4efa02ed6dc13fc307f2d87c02cf3327.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "bafde4344378c50521a2645a7e190a104176bf1d5e41288d2c51c62674f61ecd", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "5a64e0ef3d62f692eee9c9f10059a42d727125efa7d44e031bbb21c95660b701", "generation_seconds_batch": 20.066212925012223, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2288196086883545, "target_z": 0.37109383092674947, "ce": 6.316492557525635, "genuineness": 4.168440818786621, "identity": 0.32001015543937683, "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_DRAM_reg/bafde4344378c50521a2645a7e190a104176bf1d5e41288d2c51c62674f61ecd.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": "bb5b890535614fb602630a1012ae072d74d3f3efde53acab559bcfd07fb776a5", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "cc1e00c631ebea33513e9e7de372420d64031c7af19acb8a3c0878c8580f5772", "generation_seconds_batch": 22.50282762304414, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.132259368896484, "target_z": 1.164062397002319, "ce": 5.657250881195068, "genuineness": 4.061748504638672, "identity": 0.6553571224212646, "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_DRAM_reg/bb5b890535614fb602630a1012ae072d74d3f3efde53acab559bcfd07fb776a5.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "bbaa9ca061ab534a030d5e30bd45ed1458f5cf35bc41111c6d46756520144e52", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "50b00332f2568f0403319ca8454acd611533d75b73225f2ced3e1b6fcfcd5a72", "generation_seconds_batch": 20.46538573107682, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2332699298858643, "target_z": 0.37499997384185874, "ce": 6.173524379730225, "genuineness": 4.7081804275512695, "identity": 0.6206746101379395, "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_DRAM_reg/bbaa9ca061ab534a030d5e30bd45ed1458f5cf35bc41111c6d46756520144e52.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "bbe9cc4d581aa28187c4ce930d31787f38e4c20775bac0fe4fac711445a1cb0a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "5636576583e285d9861457c9e610802ada362943d10358b19f0ea66dee5b31d4", "generation_seconds_batch": 15.729167327983305, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.817627906799316, "target_z": 1.7656251471395445, "ce": 5.589737892150879, "genuineness": 3.8922953605651855, "identity": 0.23537589609622955, "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_DRAM_reg/bbe9cc4d581aa28187c4ce930d31787f38e4c20775bac0fe4fac711445a1cb0a.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "bc1362ec3fbd7cf2235fe6cb26d66c4b308a402583d1c7d4a5738890213da205", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "c9a61d45ef6dea289b1eeee5982688f9037e49da961514a6a9ee5ba4cc3f1713", "generation_seconds_batch": 22.134722555987537, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1607131958007812, "target_z": -0.5664062083104624, "ce": 6.23706579208374, "genuineness": 4.457139015197754, "identity": 0.437665194272995, "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_DRAM_reg/bc1362ec3fbd7cf2235fe6cb26d66c4b308a402583d1c7d4a5738890213da205.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "bc255110b0604de01340a82a1524ffa8076e81c9304667a53d2fa814e8934e20", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "70fede8f0b6aa059a0db0ae9b492ab900fa0b28174d44d09d9be8b32b1d6d56e", "generation_seconds_batch": 21.347048301016912, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6246719360351562, "target_z": -0.15917961372586728, "ce": 6.1894097328186035, "genuineness": 4.300237655639648, "identity": 0.5110265016555786, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/bc255110b0604de01340a82a1524ffa8076e81c9304667a53d2fa814e8934e20.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "bd9485076c1cdb21424d8aa11726f89381e7b8420ac7fcf078b7a28dc67d5ccb", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "f52da0770bd72fc597010f4420172de64c8fc5f21323eb499ff98870f918c5ba", "generation_seconds_batch": 20.516600609989837, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.048576593399048, "target_z": 0.21289064829709453, "ce": 6.268836498260498, "genuineness": 3.8546390533447266, "identity": 0.3897397518157959, "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/manual-comparison-v1/audio/vn_S_DRAM_reg/bd9485076c1cdb21424d8aa11726f89381e7b8420ac7fcf078b7a28dc67d5ccb.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "be3051e7b5a747bda19917300d2960c61566bc32bbe80eac41811acf89bb93dd", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "83ac9bd2d5f7ff8164f8be50882e96a045e95782b588120e8f09d8a3ab1d1275", "generation_seconds_batch": 21.269453841028735, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8104257583618164, "target_z": 0.0038605231167864744, "ce": 6.23706579208374, "genuineness": 4.369274139404297, "identity": 0.5118649005889893, "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_DRAM_reg/be3051e7b5a747bda19917300d2960c61566bc32bbe80eac41811acf89bb93dd.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "be3277547a367289fd24eb5a26a3b9663f15a60ea534d020af9468a979a8201a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "b90404314e7539ed6a873d4bf49c24aa2b167347851edf4761f9b6e73cb0f6e6", "generation_seconds_batch": 21.415854312945157, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.577303886413574, "target_z": 1.5546875703000045, "ce": 5.129063129425049, "genuineness": 3.9676074981689453, "identity": 0.2648729681968689, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/be3277547a367289fd24eb5a26a3b9663f15a60ea534d020af9468a979a8201a.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "be7e56554b88e5fbf3c06162acfa35cc15bb26eaa9b9ab3f5581d7760978e92b", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "314340d8c67e4a83a2c6499b265ca90d74a8e30f1f46691bc5aacff3e3fad7cc", "generation_seconds_batch": 20.146372135030106, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.893923759460449, "target_z": 0.07714851985727281, "ce": 6.3641486167907715, "genuineness": 4.7332844734191895, "identity": 0.5886650085449219, "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_DRAM_reg/be7e56554b88e5fbf3c06162acfa35cc15bb26eaa9b9ab3f5581d7760978e92b.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "bea10327b12b867c1cc29536084f1bb37f508f879654b6084af512f6b81c1b6c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "6edc62027bb00587604409543eb050e338a2d2b00457defba21c0c35db6f77d4", "generation_seconds_batch": 14.958771220990457, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.411287784576416, "target_z": 0.531250085013959, "ce": 6.046441555023193, "genuineness": 4.2374773025512695, "identity": 0.5305036902427673, "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_DRAM_reg/bea10327b12b867c1cc29536084f1bb37f508f879654b6084af512f6b81c1b6c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "bed8baa50c09a15c0e346fe7cfd3df25e19cb501ad91fdca71c614e948cdd4d5", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "827bb8b60037e1ea61a4f0657bbd757826d0bc68daac617e42a3e1c774b3a6a6", "generation_seconds_batch": 21.531054439023137, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.455792188644409, "target_z": 0.5703125604907017, "ce": 6.0146708488464355, "genuineness": 4.519899368286133, "identity": 0.6642465591430664, "hypothesis": "Ich hab die kleine Schachtel 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_DRAM_reg/bed8baa50c09a15c0e346fe7cfd3df25e19cb501ad91fdca71c614e948cdd4d5.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "bf00b7789c141797e34d261f61b08fca08fd9d582c647624b70637181f5d7ece", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "637e645ff245ac60df88ec63bfb034824a379ae6a8ecffaaf01d0e410b48a068", "generation_seconds_batch": 21.721959206974134, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2096681594848633, "target_z": -0.5234373806534806, "ce": 6.284721851348877, "genuineness": 4.469690799713135, "identity": 0.5851577520370483, "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_DRAM_reg/bf00b7789c141797e34d261f61b08fca08fd9d582c647624b70637181f5d7ece.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "bf4d66cf75df82c1a27528fa4fc2fc4d668d9de30a2b45c272e960ab6ac3b30f", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "2553f5b6b88c0750133b3341d04507d2b935091e6a7796eeec95909142f26bb9", "generation_seconds_batch": 21.09280116390437, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7940258979797363, "target_z": 0.8671875833790752, "ce": 6.109982967376709, "genuineness": 4.306513786315918, "identity": 0.48949673771858215, "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": 0.03999999999999915, "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_DRAM_reg/bf4d66cf75df82c1a27528fa4fc2fc4d668d9de30a2b45c272e960ab6ac3b30f.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "bf8bcf2b08c0dc43080883c2ceb9917fa48d2af5799ac40e276bf4b339972f58", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "aec8b7067b0c49e042e9b192aea73bb17d9e8ff4fbcf3e3e989fd4327dae93c9", "generation_seconds_batch": 21.30677841603756, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7183682918548584, "target_z": 0.8007812704360479, "ce": 6.046441555023193, "genuineness": 3.9676074981689453, "identity": 0.507075846195221, "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_DRAM_reg/bf8bcf2b08c0dc43080883c2ceb9917fa48d2af5799ac40e276bf4b339972f58.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "bfa9e4cec312e2eda1359fa081fbb98b4ea0650151c6acedb7cc5f18fcccbbcd", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "a8073b1b9d009b61f468e5d43a5da9fbe2112d6a65a79add7523033bbbf75f9e", "generation_seconds_batch": 20.56053585698828, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.03434944152832, "target_z": 1.0781247416883553, "ce": 5.019851207733154, "genuineness": 4.231201171875, "identity": 0.449387788772583, "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_DRAM_reg/bfa9e4cec312e2eda1359fa081fbb98b4ea0650151c6acedb7cc5f18fcccbbcd.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "bfcf9e60676755d554b84c6614a4fc1b10b08552bab255e7ff3026d6f3c929c2", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "f8d01f367e215502236b3fcb3d8ed27556df7b1971dbb7faef3705fc05922652", "generation_seconds_batch": 20.351392797078006, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.335630178451538, "target_z": 0.4648437720709317, "ce": 6.284721851348877, "genuineness": 4.406930446624756, "identity": 0.5659773349761963, "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_DRAM_reg/bfcf9e60676755d554b84c6614a4fc1b10b08552bab255e7ff3026d6f3c929c2.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "c0043f27ad58474f93d6f1db4c97ef6fa31ded385ad2c5fe054afc83b27a3cb5", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "91a475e675d47a923a9c3b12c69f06edd04a574dcb8ea48ac7e3dd153595b039", "generation_seconds_batch": 21.377519707079045, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.28357458114624, "target_z": 1.2968750228883736, "ce": 5.669164657592773, "genuineness": 4.344170093536377, "identity": 0.44000768661499023, "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_DRAM_reg/c0043f27ad58474f93d6f1db4c97ef6fa31ded385ad2c5fe054afc83b27a3cb5.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "c05da1c06f8b53d2fea046c703d16147b10bb189f792a7e6d55db05db9a1946a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "58c1213b0b32ffe8778001e1e5e95454c5e973404704d28e6888e6cc57545e04", "generation_seconds_batch": 16.067865638062358, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5180983543395996, "target_z": 0.6250000261581412, "ce": 5.855817794799805, "genuineness": 4.68307638168335, "identity": 0.8085808753967285, "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_DRAM_reg/c05da1c06f8b53d2fea046c703d16147b10bb189f792a7e6d55db05db9a1946a.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "c0c3d6b7a2499a49269bfc0d023210abc226645f4ca53e09ac00444b8664e138", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "f216da4dfeb509aa9a259ba1c69997ff76640dfaaedbec0ea5ba8b78b2b045e5", "generation_seconds_batch": 20.911640733946115, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0385630130767822, "target_z": 0.20410151284040373, "ce": 5.80021858215332, "genuineness": 4.570107460021973, "identity": 0.5985735654830933, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la... près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.15789473684210525, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_DRAM_reg/c0c3d6b7a2499a49269bfc0d023210abc226645f4ca53e09ac00444b8664e138.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "c117737de5403a5243223dd1f85303bb3f890e38fa66f87a1e8ec78f146581d7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "b540b766ffc4963f531f8d7c3a4f6770d72756249919d1debbd3b768fa630fbc", "generation_seconds_batch": 21.882755789905787, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.533437728881836, "target_z": -0.23925784540203704, "ce": 6.09409761428833, "genuineness": 4.6705241203308105, "identity": 0.6108077764511108, "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_DRAM_reg/c117737de5403a5243223dd1f85303bb3f890e38fa66f87a1e8ec78f146581d7.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "c1d47354209aaa3a4b5bb0c2615f7053159dfc9a10acbd04bb4884de0ac928a5", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "6ffc423cc630523f9c282d465d533df7af9b8329b7f6f89750a38fde5bda791c", "generation_seconds_batch": 21.39462256198749, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9323086738586426, "target_z": 0.1108397872443277, "ce": 6.260893821716309, "genuineness": 4.457139015197754, "identity": 0.5155823826789856, "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_DRAM_reg/c1d47354209aaa3a4b5bb0c2615f7053159dfc9a10acbd04bb4884de0ac928a5.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "c252af329e34dfbabaf45fc79b92c8eef942d696e09cede75455271838477615", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "f759187626389e413b8d04bcd42cc3d65a24c5684f8897c9a37a0b47bfdd5026", "generation_seconds_batch": 21.232470880029723, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.25997257232666, "target_z": 0.3984374591279043, "ce": 6.24500846862793, "genuineness": 4.406930446624756, "identity": 0.5365738272666931, "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_DRAM_reg/c252af329e34dfbabaf45fc79b92c8eef942d696e09cede75455271838477615.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "c280d327d13f2e757168a70e2b6976da995439dc53ed528ead9e8da670c4c9f7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "07b76f86706c1b9e1e66976bd21dc7cc3d3bdc61ee08ac758bb5f0922fde5eb8", "generation_seconds_batch": 20.709671157994308, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7220253944396973, "target_z": -0.07373038322513978, "ce": 6.141753673553467, "genuineness": 4.457139015197754, "identity": 0.7004072070121765, "hypothesis": "J'ai dans mon beau crâne. J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient?", "wer": 0.2631578947368421, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_DRAM_reg/c280d327d13f2e757168a70e2b6976da995439dc53ed528ead9e8da670c4c9f7.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "c284d08382bb2a31577f63b4a82bfa66f7af11724671443ec21f559779a1ec55", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "b0c13e9bccdd53164ea971db8e35529550bd07c71707abe287ae1db816c6fb42", "generation_seconds_batch": 20.81285622192081, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4735939502716064, "target_z": 0.5859375506813986, "ce": 6.252951145172119, "genuineness": 4.118232727050781, "identity": 0.43222594261169434, "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_DRAM_reg/c284d08382bb2a31577f63b4a82bfa66f7af11724671443ec21f559779a1ec55.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "c2fdad0ed742d3ea10efc00e25db55ea170eca679d191b183dc2cf24ed5c0d4d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "5eaba0f49fc35bbefad41e55c79aa677738085cf8f0b92eb36fa7e18cc34b65e", "generation_seconds_batch": 21.19263188203331, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9898452758789062, "target_z": 1.0390624754767426, "ce": 5.327630043029785, "genuineness": 4.193545341491699, "identity": 0.5220654606819153, "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_DRAM_reg/c2fdad0ed742d3ea10efc00e25db55ea170eca679d191b183dc2cf24ed5c0d4d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "c32a5ba9cef4a6a39194786f9694041149bd7ebf81ccca2004bc84ea67c21948", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "32661b825c2850b6ccd5d303d921dd9e9e742e30664a7244538b057eeab91089", "generation_seconds_batch": 21.26746656699106, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.176764011383057, "target_z": 1.2031250817441914, "ce": 6.221180438995361, "genuineness": 4.168440818786621, "identity": 0.5855717062950134, "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_DRAM_reg/c32a5ba9cef4a6a39194786f9694041149bd7ebf81ccca2004bc84ea67c21948.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": "c336ec7543274c7071b835ec155265ae8568031461b95a85e7422b5216471e8f", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0c4887bf8e2e2c995f57f44d8a79f40d2a76337676c8e66e4fa0bd165a01b97d", "generation_seconds_batch": 20.588639173889533, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.798476219177246, "target_z": 0.8710937262941845, "ce": 5.418970584869385, "genuineness": 3.9927117824554443, "identity": 0.507720947265625, "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_DRAM_reg/c336ec7543274c7071b835ec155265ae8568031461b95a85e7422b5216471e8f.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "c3706427d3cf774ebb70dc08e80be981f8eb2822a4fbcbe3cc1c3d2975e555fb", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "9952ad26e9c0ed6b745e79678ac655e8676df2170bf7e7027c595ca2ae3bc82f", "generation_seconds_batch": 14.773813166073523, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5937557220458984, "target_z": 0.6914061298360387, "ce": 6.133810997009277, "genuineness": 4.406930446624756, "identity": 0.4582940340042114, "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_DRAM_reg/c3706427d3cf774ebb70dc08e80be981f8eb2822a4fbcbe3cc1c3d2975e555fb.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "76f356d303adf0cce2bc8e49c7fd4094aa0a5492bc90a05a71277349ba806dce", "generation_seconds_batch": 15.754555492079817, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4544425010681152, "target_z": -0.3085936608988314, "ce": 6.284721851348877, "genuineness": 4.457139015197754, "identity": 0.5904446840286255, "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_DRAM_reg/c3977db0634952a7f9e4083e6135fbbc87199d34c7cb8c3338515d9ae38c18f4.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "c3afaf3fda8d95270bfbf45100a9710caa5cf92a973f2957d3933417146fb2b7", "bucket": "vn_S_DRAM_reg", "phase": "wording", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "f821a560a8c97e12c22f0857316b96873cd8598e5bceca1a11b74795fb124e39", "generation_seconds_batch": 14.487498109927401, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4210641384124756, "target_z": -0.3378905698226708, "ce": 6.41180419921875, "genuineness": 4.570107460021973, "identity": 0.38899970054626465, "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_DRAM_reg/c3afaf3fda8d95270bfbf45100a9710caa5cf92a973f2957d3933417146fb2b7.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "c3bd26ba671e897ddd180908fd881d1ebdef183e2b93b7b7c74232f41d92a45b", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "d9e9a6311b4d6e0892c68674f8b8dc7415258f477a0d9358105af510cab4c218", "generation_seconds_batch": 18.531556913047098, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.390385150909424, "target_z": 1.3906249640325559, "ce": 5.319687366485596, "genuineness": 3.886019229888916, "identity": 0.1661325991153717, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond er naast het olegaan. We kunnen er samen inken en kijken.", "wer": 0.2777777777777778, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_DRAM_reg/c3bd26ba671e897ddd180908fd881d1ebdef183e2b93b7b7c74232f41d92a45b.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "c41a2db40f58f4e26f17ca23c42eee2b8bbe4429e8c965e1553fa5b9a53bff55", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "4759327e79c02c8e6c1f1834270ff6fa17f850f7c5845b37ff8a590bd4ba3c22", "generation_seconds_batch": 20.547909564105794, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.656062126159668, "target_z": 0.7460938047686082, "ce": 5.530168056488037, "genuineness": 4.030367851257324, "identity": 0.3216767907142639, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/c41a2db40f58f4e26f17ca23c42eee2b8bbe4429e8c965e1553fa5b9a53bff55.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "82d7138a2e784ab443a8fc3792a24d7777d1a6c19f4735bb3ccc200450ad98d7", "generation_seconds_batch": 15.767950225970708, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.196316719055176, "target_z": -0.5351562279290684, "ce": 6.23706579208374, "genuineness": 4.331617832183838, "identity": 0.5591800212860107, "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_DRAM_reg/c493619abd8899c9b3ce96f2a00a6bcdd900e874ef21436707b986365a8437ee.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "c4988ec62c07060f455526869e206684ad037291e0338b6b6247184515ba82e0", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "32247f271e1db71ff90d70e8aae5f48e8a20b0748108568dc3947e749253892b", "generation_seconds_batch": 20.351392797078006, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9528920650482178, "target_z": 0.1289062737058155, "ce": 6.221180438995361, "genuineness": 4.2751336097717285, "identity": 0.6058238744735718, "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_DRAM_reg/c4988ec62c07060f455526869e206684ad037291e0338b6b6247184515ba82e0.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "c4a1edfd2fcfb27bc6d672e2ae843d47e9a649bf0c38a4cd05d99f1cdb87fcb7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "fc1ddb2c38d75f1061eed35681320f22aa693a6e12df02f00800a331697fd01a", "generation_seconds_batch": 14.675983360037208, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.774874210357666, "target_z": -0.02734383746628477, "ce": 6.292664527893066, "genuineness": 4.607763767242432, "identity": 0.5094861388206482, "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_DRAM_reg/c4a1edfd2fcfb27bc6d672e2ae843d47e9a649bf0c38a4cd05d99f1cdb87fcb7.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "c4ad9c9212b1571fa975d044780d983316264aed534f742be022e2f59d04dc1a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "6695bd607c8823217fe2dc9ffede21fcb81e9f37046afd2887d2761011a039e4", "generation_seconds_batch": 22.101238091941923, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5670533180236816, "target_z": 0.667968853815123, "ce": 6.2767791748046875, "genuineness": 4.381826400756836, "identity": 0.28849104046821594, "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_DRAM_reg/c4ad9c9212b1571fa975d044780d983316264aed534f742be022e2f59d04dc1a.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": "c52aedbf87ddcd04cf963dbe7e08101e90226614d04a239a814dd45afb19f1c2", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The road turns left behind the bridge.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "515b2809798500cf9f73baa8e9186a055ccabd9c3040a8791958dd8b3fcf1c23", "generation_seconds_batch": 15.32835327507928, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.497195720672607, "target_z": 1.484374905176738, "ce": 6.070269584655762, "genuineness": 4.017816066741943, "identity": 0.19535331428050995, "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_DRAM_reg/c52aedbf87ddcd04cf963dbe7e08101e90226614d04a239a814dd45afb19f1c2.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "c635de0410829ea3ba6ef0b24bca5b25984e23b65a56fb84f2de75180fbffb46", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "2647f6752f02da418ea8e752820a122c179dd716d14efc2eebee13603dfc1079", "generation_seconds_batch": 16.094509107992053, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.070828914642334, "target_z": 0.2324219906680308, "ce": 6.252951145172119, "genuineness": 3.973883628845215, "identity": 0.34138035774230957, "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_DRAM_reg/c635de0410829ea3ba6ef0b24bca5b25984e23b65a56fb84f2de75180fbffb46.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "c66d0ffbc41b406c8ac34ee4cfd42b7ba1af683123268356ed841fb47481e092", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "7924255ab3919380fb3609257519cffcdf21ad6b053edff52ef48f77d7d74b47", "generation_seconds_batch": 16.149571178946644, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8747310638427734, "target_z": 0.06030267689861544, "ce": 6.1020402908325195, "genuineness": 4.7332844734191895, "identity": 0.7519584894180298, "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_DRAM_reg/c66d0ffbc41b406c8ac34ee4cfd42b7ba1af683123268356ed841fb47481e092.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "c6e545c5c7a9f5cb190af9cdfb98b089e5463dcf8ed0094f9371618725fe5735", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "512f9a04bf1d2b41882b0522ff0996a5298b79d4a498eb718d35e7ae433048d9", "generation_seconds_batch": 23.789757214020938, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4922711849212646, "target_z": -0.27539060905988266, "ce": 6.09409761428833, "genuineness": 4.099404335021973, "identity": 0.5230917930603027, "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_DRAM_reg/c6e545c5c7a9f5cb190af9cdfb98b089e5463dcf8ed0094f9371618725fe5735.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "c78fc03ab60470f26d8ddec7fd085eee1ce03f8877efc193361e4aed3f4b0157", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "7a712a6e4e00d7fdd04462fbbb4d5273d4eee2e064fd829f4125ef8adc559c0e", "generation_seconds_batch": 21.348820040933788, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.89638614654541, "target_z": 0.9570313816081482, "ce": 5.220403671264648, "genuineness": 3.9676074981689453, "identity": 0.40818238258361816, "hypothesis": "De deuren staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "wer": 0.05263157894736842, "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_DRAM_reg/c78fc03ab60470f26d8ddec7fd085eee1ce03f8877efc193361e4aed3f4b0157.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "460d0ece05e7f6e62ba592231361b19afd019fe6ed9e1a8f0ee5488eb2cc0c7e", "generation_seconds_batch": 16.21426491998136, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.071704387664795, "target_z": -0.6445311592639476, "ce": 6.173524379730225, "genuineness": 4.193545341491699, "identity": 0.7209322452545166, "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_DRAM_reg/c82131bfaf89850d394bf3a21117deb94436b65306054c3e1cbac0c65e2f913d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "c85e068ed7932103c5cbb1262a134e652c7a6d14685a5dd68b72cc1fd99f216e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "c56cb57b28efcd8dd9a22214c17334101e40cd343f30640cf32276faa4778b4f", "generation_seconds_batch": 22.961062068934552, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.245271682739258, "target_z": -0.4921874002720865, "ce": 6.054384231567383, "genuineness": 4.337893962860107, "identity": 0.3409004807472229, "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": 0.03999999999999915, "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_DRAM_reg/c85e068ed7932103c5cbb1262a134e652c7a6d14685a5dd68b72cc1fd99f216e.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "7d82425c4a54f73624b6552f31ae3b07300267ef16751603a6b5d15ad523fa8a", "generation_seconds_batch": 15.135061737964861, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5223116874694824, "target_z": -0.24902341195494018, "ce": 6.38003396987915, "genuineness": 4.406930446624756, "identity": 0.29657846689224243, "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_DRAM_reg/c8907269f81c0a33f09ba3b9d9f76b6acb2a8250e6881dbc11786cea01719e9d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "c890f73264f9afd5328ea036841cf3ddfda2af14136d5c529665fd9277fc6467", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "f29636af72c340ec462c9dda95307756ee5ac1e424dd78a64aa7cb02031c16b0", "generation_seconds_batch": 20.68309138098266, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.876121997833252, "target_z": 0.061523529666575774, "ce": 6.133810997009277, "genuineness": 4.287685871124268, "identity": 0.6693230867385864, "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_DRAM_reg/c890f73264f9afd5328ea036841cf3ddfda2af14136d5c529665fd9277fc6467.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "c8a30b33615033c44412cf3f60a3f54df7ef3bb48f5129a8ca899046d112380e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "b5eb5a1aaf0e414366cf36c423123cb5ef1bb35a4471f8e2124bab91d9c3b652", "generation_seconds_batch": 19.98608150705695, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.696991443634033, "target_z": -0.09570322186686678, "ce": 6.173524379730225, "genuineness": 4.557555198669434, "identity": 0.7903459072113037, "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_DRAM_reg/c8a30b33615033c44412cf3f60a3f54df7ef3bb48f5129a8ca899046d112380e.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "c97724ab76a9ceca72adc7f9df3564fb522958810c5480cfaaecd1183e0464e8", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "97e906cc2a932cd5a47bd0e9c5826b7ebd73f0310c38ca95773e8def1ab43bb3", "generation_seconds_batch": 20.96969827800058, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0051846504211426, "target_z": 0.17480460391656433, "ce": 6.300607204437256, "genuineness": 4.432034492492676, "identity": 0.6660599112510681, "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_DRAM_reg/c97724ab76a9ceca72adc7f9df3564fb522958810c5480cfaaecd1183e0464e8.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": "c99dbc40955d568b54512e476276f4938053d00546d9b56f15676b6cbc303057", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Someone moved the little blue chair.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "45348d15cc007cf517cadf8f8fe0e55e7731cf24dcfb972c6cdae4ff6da762d9", "generation_seconds_batch": 15.313343288027681, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.069953441619873, "target_z": 1.1093751406000092, "ce": 5.998785495758057, "genuineness": 4.068024158477783, "identity": 0.39151531457901, "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_DRAM_reg/c99dbc40955d568b54512e476276f4938053d00546d9b56f15676b6cbc303057.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "ca18c75d3cf4fbe60e1c5dbb141f3dcc2a3a4adabeefcb89ef35e4bbada341e9", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "85f954f638289eaec911b4a1358da0add6883ec62a87a595befa589b8f812be8", "generation_seconds_batch": 20.05379652197007, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.372583389282227, "target_z": 1.3749999738418588, "ce": 5.557967185974121, "genuineness": 4.093128204345703, "identity": 0.6573218107223511, "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": 0.03999999999999915, "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_DRAM_reg/ca18c75d3cf4fbe60e1c5dbb141f3dcc2a3a4adabeefcb89ef35e4bbada341e9.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "ca67cc2bd950ada8941a41de385f05b3509d927958360d08557430a2c98a42dc", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "843e1f86702f36b805e5c0573905f1b8ca5773f5518745ae0978987897b39f71", "generation_seconds_batch": 16.19132740003988, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0352251529693604, "target_z": 0.2011718010215068, "ce": 6.078212261199951, "genuineness": 4.3127899169921875, "identity": 0.3239138424396515, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/ca67cc2bd950ada8941a41de385f05b3509d927958360d08557430a2c98a42dc.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": "cafd89b497a2975a574bfa018e525e99cbff999f8e698fba076f9e89c484291c", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "ccf14b7cc2c1da8966cabdbc7eedbc1f4dacbf7034324e47839018d34dfb8774", "generation_seconds_batch": 14.658118849969469, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4246389865875244, "target_z": 0.5429687230244169, "ce": 5.6532793045043945, "genuineness": 4.507347106933594, "identity": 0.3596712052822113, "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_DRAM_reg/cafd89b497a2975a574bfa018e525e99cbff999f8e698fba076f9e89c484291c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "cb2730ff45806365681377ede98f6b79de455ee8440f736ce01abd93f9c5b523", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "dc07dcffdc796715a62a5fa7173c6b9bf991f1aaa725b5ff4031d18670a58aa0", "generation_seconds_batch": 19.797749129938893, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7139177322387695, "target_z": 0.7968749182558086, "ce": 5.804190158843994, "genuineness": 4.2249250411987305, "identity": 0.5240412950515747, "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": 0.03999999999999915, "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_DRAM_reg/cb2730ff45806365681377ede98f6b79de455ee8440f736ce01abd93f9c5b523.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "cb428bcd54589ee23fbedb885a76315d7fb75af71bca9845e404c70e5b54807d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "1e58a0d9ec94a41a9eb92603d46c8dde1c7a4ad4d67a8ea21d5274523e844678", "generation_seconds_batch": 21.65460466197692, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.705016851425171, "target_z": 0.7890624231604602, "ce": 5.198561668395996, "genuineness": 4.111956596374512, "identity": 0.43181049823760986, "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_DRAM_reg/cb428bcd54589ee23fbedb885a76315d7fb75af71bca9845e404c70e5b54807d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "cb970317ea028daafe9903e32985f2f48bdd47847016d186313f1ecc4f91b028", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "e8a28b636b658b5f4ce742371f612bab3f2dd92b40da044a27710e1874e7bfeb", "generation_seconds_batch": 15.734572071931325, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8193788528442383, "target_z": 0.011718847275587738, "ce": 6.260893821716309, "genuineness": 4.8588056564331055, "identity": 0.5870126485824585, "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_DRAM_reg/cb970317ea028daafe9903e32985f2f48bdd47847016d186313f1ecc4f91b028.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "cbbec16ae9e6d5c5cd247fb3a1b63464f4ab0eb571bc7a01d72034a6114f1769", "bucket": "vn_S_DRAM_reg", "phase": "wording", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "80f99f5cbe1bcd0498e726fd34bf5262080c516adcc272ce80ba1af5c34eed6a", "generation_seconds_batch": 16.625647816923447, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2474968433380127, "target_z": -0.4902343288145319, "ce": 6.1020402908325195, "genuineness": 4.444586753845215, "identity": 0.7646653652191162, "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_DRAM_reg/cbbec16ae9e6d5c5cd247fb3a1b63464f4ab0eb571bc7a01d72034a6114f1769.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "cbebf04af8765d7bde5a302c055c4a87ad53d7b92bc64d8c017882bab7b38b7a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "009e6d4e8b3bcce6b924ff6febd6cdd4f17cc580a47f34db8e19abcbbda4c8fc", "generation_seconds_batch": 14.236820242018439, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.36543345451355, "target_z": -0.3867188211174465, "ce": 6.268836498260498, "genuineness": 4.118232727050781, "identity": 0.5938649773597717, "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_DRAM_reg/cbebf04af8765d7bde5a302c055c4a87ad53d7b92bc64d8c017882bab7b38b7a.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "cc3546c61e0ffe4313fa95578455af3d6307f980ddd4d6c552d479b7f3de8a16", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "fd199d03f7e5480009edad63cd6d4a733557ce6c011fbad0c10e9284a63a6724", "generation_seconds_batch": 19.80487561505288, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2288196086883545, "target_z": 0.37109383092674947, "ce": 6.046441555023193, "genuineness": 4.519899368286133, "identity": 0.72690749168396, "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_DRAM_reg/cc3546c61e0ffe4313fa95578455af3d6307f980ddd4d6c552d479b7f3de8a16.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "cc9d7168bfdb645c2ba549c7ba96e28421fcc2c8b1d23292bc3251abb7e222c0", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "34d14b20ffb704034542794f04e076e972a02f65a13309653fcf783e06862520", "generation_seconds_batch": 21.721959206974134, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8193788528442383, "target_z": 0.011718847275587738, "ce": 6.268836498260498, "genuineness": 4.469690799713135, "identity": 0.6249701380729675, "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_DRAM_reg/cc9d7168bfdb645c2ba549c7ba96e28421fcc2c8b1d23292bc3251abb7e222c0.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "cd54f0a7ffafdbdd4d254e15d0f94d4c7648d014c7c06bdf40d6c66581242c9e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "e2e8c13281cb769e609c685f390e6d502b19248a9f1ffe5dc7ef3e89859672ae", "generation_seconds_batch": 20.621511646080762, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9851577281951904, "target_z": 0.15722654226831265, "ce": 6.165581703186035, "genuineness": 4.080576419830322, "identity": 0.4247514605522156, "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_DRAM_reg/cd54f0a7ffafdbdd4d254e15d0f94d4c7648d014c7c06bdf40d6c66581242c9e.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "cd6d300601a8fa39cc5f13a2bee4b9fe4d7e82b8aa2534b7813351603a795537", "bucket": "vn_S_DRAM_reg", "phase": "wording", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "618051c38217ab620cf208428143a1b86bb35601ca74a740d3a4ae5affc7a62c", "generation_seconds_batch": 16.127483104006387, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2096681594848633, "target_z": -0.5234373806534806, "ce": 6.252951145172119, "genuineness": 4.557555198669434, "identity": 0.6451406478881836, "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_DRAM_reg/cd6d300601a8fa39cc5f13a2bee4b9fe4d7e82b8aa2534b7813351603a795537.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "ce9857551d738be079618c1fd31565efed0e40e011e022b4382d0bcdbd0a9e56", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "e7fc579e54c3dae0a49cd9669427992ae5cb52c11f758cfd41248f7751ea753b", "generation_seconds_batch": 20.740291732014157, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3979363441467285, "target_z": 0.5195312377383713, "ce": 5.911416053771973, "genuineness": 4.325342178344727, "identity": 0.3651208281517029, "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_DRAM_reg/ce9857551d738be079618c1fd31565efed0e40e011e022b4382d0bcdbd0a9e56.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "cf162930d2fc82dd0a4dc3ebd57239f843718060ef45f59f2ea66c0e9b1ca96c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "b5bcabc63d3557b330e863b41870d3959dc862c6c9345505fb47f1d793c9216e", "generation_seconds_batch": 20.66261321795173, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2577474117279053, "target_z": 0.39648438767034966, "ce": 6.221180438995361, "genuineness": 4.419482707977295, "identity": 0.4719308018684387, "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_DRAM_reg/cf162930d2fc82dd0a4dc3ebd57239f843718060ef45f59f2ea66c0e9b1ca96c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "cf41fad0949d7e08a464123106162dfe34b2b6db0fc3f7cc2cc5ed0d67fe4462", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "af48a4654f4cb4667e027118729d61e67e31a44813d8dba3d8d320ff08f94e49", "generation_seconds_batch": 16.032449984923005, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9896082878112793, "target_z": 0.16113289444855186, "ce": 6.332377910614014, "genuineness": 4.507347106933594, "identity": 0.6102526187896729, "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_DRAM_reg/cf41fad0949d7e08a464123106162dfe34b2b6db0fc3f7cc2cc5ed0d67fe4462.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "cf58ed984319aba46b94573001d8948d287f4e73d0c1d094780c7036cd388f33", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "3c8d3028c86ad955ba64e8c5f44bd0f8a3c2da223529ae1a125f426c885d1c42", "generation_seconds_batch": 21.256687902030535, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.146486282348633, "target_z": 0.2988280943459282, "ce": 6.038498878479004, "genuineness": 4.557555198669434, "identity": 0.540988028049469, "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_DRAM_reg/cf58ed984319aba46b94573001d8948d287f4e73d0c1d094780c7036cd388f33.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "d87d6937eb1dd82cfefda315b23a62686b4f73ac013910aac498d3c2fbb1b32b", "generation_seconds_batch": 15.704136588959955, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.363208293914795, "target_z": -0.3886718925750011, "ce": 6.3641486167907715, "genuineness": 4.7332844734191895, "identity": 0.6642399430274963, "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_DRAM_reg/cf68e6506caaaad7e7371e72de9d5417fef89211300487ff83acd2484eaab510.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "d01a23b33401bca143c145bb53f323afc981c596b5e8099deb1cb55786bb34e6", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "44afb6ce72ef6aa245eb9845d800d3ef366ab0c2237ee633154d48e7fcf3c916", "generation_seconds_batch": 16.51156911405269, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.795457601547241, "target_z": -0.009277351004796985, "ce": 6.459460258483887, "genuineness": 4.507347106933594, "identity": 0.4409146308898926, "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_DRAM_reg/d01a23b33401bca143c145bb53f323afc981c596b5e8099deb1cb55786bb34e6.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": "d0ebc2832d0af566dcf0376ed1bcbe0fe822793e50b1f907f938a9df7ba332a4", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3e84105c7accd1880bab1ae5ec7f54d54193052f9646b40283a7aaddda7856d3", "generation_seconds_batch": 21.12764885101933, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.455792188644409, "target_z": 0.5703125604907017, "ce": 5.9431867599487305, "genuineness": 4.394378662109375, "identity": 0.40402889251708984, "hypothesis": "Iemand heeft het kleine blauwe stoeltje verplaatst. Het stond hier naast de hoge boekenkast. We zetten het dadel lunch terug.", "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_DRAM_reg/d0ebc2832d0af566dcf0376ed1bcbe0fe822793e50b1f907f938a9df7ba332a4.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "682b69efa87244d8dc4af6f773bf9f49d3cbfc4e151a8725c4b8be55043b9962", "generation_seconds_batch": 15.367575297947042, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2786498069763184, "target_z": -0.462890700613377, "ce": 6.38797664642334, "genuineness": 4.18099308013916, "identity": 0.316811740398407, "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_DRAM_reg/d11caac31fa3732e2517803823f1b039213fb31e9438619fd83d8b232c894b7d.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": "d12516097f2d9089ea40cd19bd759bcfaeb1d2dcca9e2e7e2fd8205a21c4c116", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0777a03c5983a0981a3045d650a87fcddea520997bbfb14acdd934c018a9233a", "generation_seconds_batch": 21.033840715070255, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2555222511291504, "target_z": 0.394531316212795, "ce": 5.395142555236816, "genuineness": 4.21237325668335, "identity": 0.4546257555484772, "hypothesis": "Ik schreef het nummer op de envelop. Lees het goed en controleer de laatste twee cijfers. Wel 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_DRAM_reg/d12516097f2d9089ea40cd19bd759bcfaeb1d2dcca9e2e7e2fd8205a21c4c116.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "d12dfe4a0403dfbf43e9c26560b9390d5327dffa19f82b927ccec60286bd4f3c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "266a931185090dab1c04c83395c066aa23c185f044ea75edc2bc7c3f9f65c4c4", "generation_seconds_batch": 18.86060091597028, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5448009967803955, "target_z": 0.6484375114441868, "ce": 5.7128496170043945, "genuineness": 4.319066047668457, "identity": 0.7188007235527039, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt. Dis-moi quand tu es prêt. Dis-moi quand tu es prêt.", "wer": 0.6666666666666666, "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_DRAM_reg/d12dfe4a0403dfbf43e9c26560b9390d5327dffa19f82b927ccec60286bd4f3c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "d134e19b584690453953dc8d5e5ac7d01a9bbbedf3b1854246ddd93822352925", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "4c9e6db7b7d8fd3ed9ad2a641fe5ce9d09951de34176bfbb512a55e15356312a", "generation_seconds_batch": 21.14436552999541, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.308927536010742, "target_z": 0.44140628678488614, "ce": 6.300607204437256, "genuineness": 4.331617832183838, "identity": 0.45594459772109985, "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_DRAM_reg/d134e19b584690453953dc8d5e5ac7d01a9bbbedf3b1854246ddd93822352925.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "d13d6ae76808f5ec41d32f8e4a33b4914b184061acd2c7a330575d51bb429036", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "102be8f62fbdbe47da102616b807f5eeaaee2727f7396816a14031038ded50de", "generation_seconds_batch": 21.304134557954967, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.514997482299805, "target_z": 1.499999895367435, "ce": 5.696964263916016, "genuineness": 3.886019229888916, "identity": 0.5355145931243896, "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_DRAM_reg/d13d6ae76808f5ec41d32f8e4a33b4914b184061acd2c7a330575d51bb429036.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "d167cfa7a0f3e72e7dce276f60e2200115f35e5a44ac6dc28105ed158de83fca", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "c26f96f9aad60053629e75c45929dd19a42bca7266000411270a34f591b5f7d1", "generation_seconds_batch": 15.315205003018491, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0118603706359863, "target_z": 0.18066402755435818, "ce": 6.205295085906982, "genuineness": 4.7332844734191895, "identity": 0.7188777923583984, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Inima, j'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.4444444444444444, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_DRAM_reg/d167cfa7a0f3e72e7dce276f60e2200115f35e5a44ac6dc28105ed158de83fca.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "d16ac7459791b4a5851bb0c1c971f28b7901ce5ad1934f7f1aba223f5d67470c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "4ac51b18cebe2906f7df55be1c2f619213c94ede6314c8986bf8b61d049b3aba", "generation_seconds_batch": 21.195769233978353, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7139177322387695, "target_z": 0.7968749182558086, "ce": 5.808161735534668, "genuineness": 4.168440818786621, "identity": 0.5821996927261353, "hypothesis": "Ich hab die kleine Schachtel 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_DRAM_reg/d16ac7459791b4a5851bb0c1c971f28b7901ce5ad1934f7f1aba223f5d67470c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "d173209f341406e9ce6f222900bd16bec7ec1f93ed6eb20801b844f221c591d0", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "c46b24bbedb8478c3604019261ccc6977c767c8844b84a00b900549336b7e450", "generation_seconds_batch": 15.519081822014414, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.820728302001953, "target_z": 0.8906248593999908, "ce": 6.157639026641846, "genuineness": 4.293961524963379, "identity": 0.32900485396385193, "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_DRAM_reg/d173209f341406e9ce6f222900bd16bec7ec1f93ed6eb20801b844f221c591d0.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "d17c818085efda43c79deaf5228827793d4cd6cfe7ac73cbaac674cb12b07bd4", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "7eb38041e3e0ae331b94bb30f2b7354dccc9ece23aaca0f592b8b53c7741481c", "generation_seconds_batch": 16.297408030950464, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.218568801879883, "target_z": -0.515625094823262, "ce": 6.09409761428833, "genuineness": 4.532451152801514, "identity": 0.5658465623855591, "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_DRAM_reg/d17c818085efda43c79deaf5228827793d4cd6cfe7ac73cbaac674cb12b07bd4.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "d1a130bb54ab4d0a79ab28f73f19bf9bf952ef189431c14e3661890b5c81c321", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "09ec31157d7e0f72df5ac4f873545a7584f52e1793b3feec8a9cf01af152ec18", "generation_seconds_batch": 15.227305427892134, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0641531944274902, "target_z": 0.22656256703023692, "ce": 5.784333229064941, "genuineness": 4.482243061065674, "identity": 0.5890229940414429, "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_DRAM_reg/d1a130bb54ab4d0a79ab28f73f19bf9bf952ef189431c14e3661890b5c81c321.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "d1db9165a4d681dc9fc98ff842da2feeb124b230af6669f14bccd9b604bbef33", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "c9a26d1f4447cc84c398234212c6bb7468c40b6e850658c87ef6df48477cc42a", "generation_seconds_batch": 19.513697082991712, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0752792358398438, "target_z": 0.2363281335831401, "ce": 5.9431867599487305, "genuineness": 4.808597087860107, "identity": 0.6566584706306458, "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_DRAM_reg/d1db9165a4d681dc9fc98ff842da2feeb124b230af6669f14bccd9b604bbef33.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "d1f30e027afa1734a0569e8779df0372ab6e70fe3512dba3f1cf4b5cefb48e2f", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "12d47038755b8857acc6fa6abb664742a5fa60f03c9ef8bdacaa477723c2f559", "generation_seconds_batch": 15.974917615996674, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.705016851425171, "target_z": 0.7890624231604602, "ce": 5.77639102935791, "genuineness": 4.331617832183838, "identity": 0.44021254777908325, "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_DRAM_reg/d1f30e027afa1734a0569e8779df0372ab6e70fe3512dba3f1cf4b5cefb48e2f.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "4f465655f7184aff4a1a9ea7745cb60840f0fde974f7fc21e01135edfe2f33fd", "generation_seconds_batch": 16.395268750027753, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8611016273498535, "target_z": 0.048339826481539575, "ce": 6.3641486167907715, "genuineness": 4.20609712600708, "identity": 0.5264882445335388, "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_DRAM_reg/d2319f0ba4d4a13f099c61fd2524d2cd47bb09bca092815ed8a29e7f5b0c527c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "d276369270312f605b531b51261bfd7c9d718b5b88e32fd7f48a8a1622a1da6b", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "09df1a08f11f30184185a50d55b0d06a3b1a5d46a9df3ca0f80b1797e72816c5", "generation_seconds_batch": 21.33135246799793, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5893054008483887, "target_z": 0.6874999869209294, "ce": 6.022613525390625, "genuineness": 4.7332844734191895, "identity": 0.4383794665336609, "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_DRAM_reg/d276369270312f605b531b51261bfd7c9d718b5b88e32fd7f48a8a1622a1da6b.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "d2ce67c385ce70ec1dd19d0c4b9679b1ea14854b8e60d30b4a120650ecc53020", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "55ec77959876a8240d5c7f0d79621a8b5fa579d18127866f86f0cbaa0e91f244", "generation_seconds_batch": 20.096152001991868, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9840450286865234, "target_z": 0.15624990190697036, "ce": 6.157639026641846, "genuineness": 4.808597087860107, "identity": 0.5006322264671326, "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": 0.03999999999999915, "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_DRAM_reg/d2ce67c385ce70ec1dd19d0c4b9679b1ea14854b8e60d30b4a120650ecc53020.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "d2f75583a5d403fab9a394696dac9cbb997f234c38d009d3eec2863f4e14992a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "168ce6fbac6116942731c8eafb41096bbb49c6db531f40e484b3b60f173cd4ec", "generation_seconds_batch": 21.278364436002448, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 5.049050807952881, "target_z": 1.9687500196186059, "ce": 5.5738525390625, "genuineness": 3.5596652030944824, "identity": 0.46773865818977356, "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_DRAM_reg/d2f75583a5d403fab9a394696dac9cbb997f234c38d009d3eec2863f4e14992a.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "d305b32391e82c7c8e1687007b6252b8bea49df8f81e7732f10bc8adad621141", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "36f6db04e8e478c51ab6e4202436639f8134d63419b662c86c725e7509021908", "generation_seconds_batch": 21.03179468109738, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.018299102783203, "target_z": -0.6914061298360387, "ce": 6.348263263702393, "genuineness": 4.595211505889893, "identity": 0.6305902004241943, "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_DRAM_reg/d305b32391e82c7c8e1687007b6252b8bea49df8f81e7732f10bc8adad621141.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "d32fac6aa4a6bb0a88255206cc3016c559efa06abb2dff14dd8f073f968c1903", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "58d34af9b64841ba46be3f1e6e05f17888db83ad5210b4612ea95bf126c65e68", "generation_seconds_batch": 22.21201195497997, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.390385150909424, "target_z": 1.3906249640325559, "ce": 5.462655067443848, "genuineness": 4.074300289154053, "identity": 0.33365362882614136, "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_DRAM_reg/d32fac6aa4a6bb0a88255206cc3016c559efa06abb2dff14dd8f073f968c1903.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "d38df10899bc32d451e608a6c433ca5727cb349ccddc73a209497b69ec9e920b", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "4085d932a575cb8fda7a5cfac054d9ec3b58bf18ce176b6e748f9008f98937ba", "generation_seconds_batch": 21.856821782072075, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.138460874557495, "target_z": -0.5859375506813986, "ce": 6.149696350097656, "genuineness": 4.425758361816406, "identity": 0.6282855868339539, "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_DRAM_reg/d38df10899bc32d451e608a6c433ca5727cb349ccddc73a209497b69ec9e920b.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "85ddba9c142734cc254fe0b56cc4081f986bd5b2b2b4ae12f8aae59609aff23f", "generation_seconds_batch": 16.003403531969525, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4010369777679443, "target_z": -0.35546884073605245, "ce": 5.990842819213867, "genuineness": 4.6203155517578125, "identity": 0.7175665497779846, "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_DRAM_reg/d3fab0ead1fc9dcf899f59bddb0c9d4b54373ba1829b6dc49f77fc4027ce1740.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": "d40acdbaa4692c2e2afe5c88b77325553ff5b41ea4d30388a442a1f034cbf391", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "8e96dda5d7e3af9fb14119e0a8fb4d887a68723a28216191a2c931405b8ca6e8", "generation_seconds_batch": 15.764855118002743, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7895753383636475, "target_z": 0.8632812311988359, "ce": 6.046441555023193, "genuineness": 4.344170093536377, "identity": 0.5304845571517944, "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_DRAM_reg/d40acdbaa4692c2e2afe5c88b77325553ff5b41ea4d30388a442a1f034cbf391.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "d42cb641004955cecd6c52e4c540b6e405e133b9c7d47f42103625e6f8ef2300", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "a03b9b88635b2172a9e448016b368a91ac053d30e9a345f258facb83317fddbd", "generation_seconds_batch": 20.518998953979462, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.497195720672607, "target_z": 1.484374905176738, "ce": 5.174733638763428, "genuineness": 3.7542223930358887, "identity": 0.20434188842773438, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude graam. We kunnen er 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_DRAM_reg/d42cb641004955cecd6c52e4c540b6e405e133b9c7d47f42103625e6f8ef2300.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "d48c761aa0c49a5832d2205a672136434a3ac8dccfbeeeffd2870893ff888610", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "459f93baa58b2d60cc88159468f78836099670ccfa42947d6ea768ac325b29e7", "generation_seconds_batch": 21.255157538922504, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.319178104400635, "target_z": 1.3281250032697676, "ce": 5.677107334136963, "genuineness": 4.036643981933594, "identity": 0.3919233977794647, "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_DRAM_reg/d48c761aa0c49a5832d2205a672136434a3ac8dccfbeeeffd2870893ff888610.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "d4fabd2258198a12b72d8c5bdc1348bd543c5d0243e6c5fa8e2e058a25201cf3", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "106a132768a6f0ff286069d664ebd9068955fe7c8177ffc82aa7f8cfe2778741", "generation_seconds_batch": 19.886398899019696, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0040721893310547, "target_z": 0.17382817282035196, "ce": 6.372091293334961, "genuineness": 4.494794845581055, "identity": 0.5931510925292969, "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_DRAM_reg/d4fabd2258198a12b72d8c5bdc1348bd543c5d0243e6c5fa8e2e058a25201cf3.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "d624a7beb0d2b2699d541f487d4cf104f8abe24f85b855f2558fb1520c140bab", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "ef47042170eaf8fcd011dbab40b5c1d49d5eab985e9b28e0a3fd7a488b7ec4d2", "generation_seconds_batch": 15.23129603289999, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.54170036315918, "target_z": 1.5234375899186106, "ce": 5.601652145385742, "genuineness": 3.9927117824554443, "identity": 0.3291057050228119, "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_DRAM_reg/d624a7beb0d2b2699d541f487d4cf104f8abe24f85b855f2558fb1520c140bab.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "d64a311d956f0af078d27d21483a342d74de564861248289bc9c43f9fe411f4d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "f87aa839866731d57b2491c6d7ca7dbb32be93ca66f044c55d6b8d33e4e663ff", "generation_seconds_batch": 21.415854312945157, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.390385150909424, "target_z": 1.3906249640325559, "ce": 5.518253803253174, "genuineness": 3.9299514293670654, "identity": 0.42981576919555664, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/d64a311d956f0af078d27d21483a342d74de564861248289bc9c43f9fe411f4d.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": "d663e407e1cb8e678772d508b3d28f498c036ba01d34611bffd0925e3a7046a9", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "40668a8fd53f79dc1420648cd0359821cb9a149b6b42b6f5026810c6bf862217", "generation_seconds_batch": 14.893597115995362, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7628726959228516, "target_z": 0.8398437459127904, "ce": 5.720792293548584, "genuineness": 4.507347106933594, "identity": 0.5978893041610718, "hypothesis": "J'ai écrit le numéro sur l'enveloppe, lue-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_DRAM_reg/d663e407e1cb8e678772d508b3d28f498c036ba01d34611bffd0925e3a7046a9.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": "d6911fa9c17fd0c8982a9aaacefb795297edd3f702a18d236b073b34cb43ed00", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e3d2d6b1ba64ba6e67e16aa0ae1525418f16d3b1af5e7109a925e480a8c9b814", "generation_seconds_batch": 21.114120552898385, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.28357458114624, "target_z": 1.2968750228883736, "ce": 5.903473377227783, "genuineness": 4.394378662109375, "identity": 0.42499619722366333, "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_DRAM_reg/d6911fa9c17fd0c8982a9aaacefb795297edd3f702a18d236b073b34cb43ed00.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "d6b9ec46be916b117d100c1cdeb0ce573b8a70fc17d285901ae1a897d45a4378", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "c1f973d644d4fc0691021cc2f6c30013432f925816c039873ae78cdc9f80c8e7", "generation_seconds_batch": 16.656995277036913, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.124234199523926, "target_z": 0.2792969612401219, "ce": 5.982900142669678, "genuineness": 4.545003414154053, "identity": 0.6532811522483826, "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_DRAM_reg/d6b9ec46be916b117d100c1cdeb0ce573b8a70fc17d285901ae1a897d45a4378.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": "d75870067d5648ea31a10b8320ab822fd9d740ad31ea33ff2a0971783a774c93", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b52342d4493d9f7757d01d3f048e16e3680c44e96464e56c28afe49558bfad0d", "generation_seconds_batch": 21.321423097047955, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.753971815109253, "target_z": 0.8320312508174419, "ce": 5.792276382446289, "genuineness": 4.300237655639648, "identity": 0.5089907050132751, "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_DRAM_reg/d75870067d5648ea31a10b8320ab822fd9d740ad31ea33ff2a0971783a774c93.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "d79a0c4abcedbccffead8ca48403799e4c9fbc30c8e263083f538b96f525b4cc", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "253020789a64ec7e59bc11bf8cf5b8bb345d2a5fb955f045af4da930643e404c", "generation_seconds_batch": 22.68269412498921, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8900294303894043, "target_z": 0.07373038322513978, "ce": 6.070269584655762, "genuineness": 4.557555198669434, "identity": 0.5328382849693298, "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_DRAM_reg/d79a0c4abcedbccffead8ca48403799e4c9fbc30c8e263083f538b96f525b4cc.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "d79d76b14387e28723290672076e29c75b63793521b37eca8b03b5bec4429bba", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "d02b0ac271235bd42e563f6d493d0e30c207aa289427e81ac7882266eaf7a388", "generation_seconds_batch": 17.676805621013045, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9631426334381104, "target_z": 1.015624990190697, "ce": 5.418970584869385, "genuineness": 4.20609712600708, "identity": 0.5072349309921265, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/d79d76b14387e28723290672076e29c75b63793521b37eca8b03b5bec4429bba.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "d7bc78cdca097fed71706ab9c102bfd25ffca427a120e0c6bff1f14999c33fef", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "f414387d62aa68b86780f8ae47f94539eae2ed1dc592e7110b61a50b19930a29", "generation_seconds_batch": 19.26112615596503, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5479016304016113, "target_z": -0.22656256703023692, "ce": 5.967014789581299, "genuineness": 4.6956281661987305, "identity": 0.46491581201553345, "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_DRAM_reg/d7bc78cdca097fed71706ab9c102bfd25ffca427a120e0c6bff1f14999c33fef.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "d7d99da12ba7fc2289bb9c0aaf2a9649aa21228a1f72ae3f21afe51bd0a53085", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "cb19dc084dae7b783af11a36897083eb7ccbaf2a968321e7030b4f565adf065d", "generation_seconds_batch": 20.50148458406329, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7673230171203613, "target_z": 0.8437498888278997, "ce": 5.816103935241699, "genuineness": 4.356721878051758, "identity": 0.6536887884140015, "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_DRAM_reg/d7d99da12ba7fc2289bb9c0aaf2a9649aa21228a1f72ae3f21afe51bd0a53085.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "d85a6cf69a064e758dd87e8d28e1fd84a527d649e81eeb82360c9239f79b9b8d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "98084b9f2f0db0fb469e63467362e09df6e1ac0b83a5d4a7ec72e47eff9257db", "generation_seconds_batch": 23.309548681019805, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3520822525024414, "target_z": -0.3984374591279043, "ce": 5.855817794799805, "genuineness": 4.18099308013916, "identity": 0.5952210426330566, "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_DRAM_reg/d85a6cf69a064e758dd87e8d28e1fd84a527d649e81eeb82360c9239f79b9b8d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "d863beaaba5ce4c4ec33e57a6484c5aba358a4e7afb7f082a451ea29db5a3fe7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "c68124149d1a4e88bc7564e2148db323b44b2d1a8e857b8a8492cf2f2e59af4c", "generation_seconds_batch": 20.9530884999549, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.123358726501465, "target_z": 1.1562501111721002, "ce": 5.335572719573975, "genuineness": 4.005263805389404, "identity": 0.7957922220230103, "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_DRAM_reg/d863beaaba5ce4c4ec33e57a6484c5aba358a4e7afb7f082a451ea29db5a3fe7.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "d9419c8fd3b53451df853da5b4202e3b55711a653193991477eea1725c143c26", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "01bc81ef9614bcfadc691163de186c3efef946d44abd783dc3baf56fdcd0c060", "generation_seconds_batch": 20.86922641703859, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.372583389282227, "target_z": 1.3749999738418588, "ce": 5.200547218322754, "genuineness": 4.080576419830322, "identity": 0.615153431892395, "hypothesis": "J'ai trouvé la petite boîte, elle était prêt, toute la vieille fenêtre, nous pouvons regarder ensemble ce qu'elle contient.", "wer": 0.10526315789473684, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_DRAM_reg/d9419c8fd3b53451df853da5b4202e3b55711a653193991477eea1725c143c26.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "d9991e08e65d2c17bcbf7bae88f3749f412f7cd370e6e5d709f935e6bfef9053", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "876657e90ca753485c34882ee3dfd55ca7900dbbd4f43f466616ba3eceae11cd", "generation_seconds_batch": 17.676805621013045, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.114457607269287, "target_z": 1.1484374068116219, "ce": 5.230332374572754, "genuineness": 4.13078498840332, "identity": 0.33334529399871826, "hypothesis": "La puerta sigue abierta. he dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/d9991e08e65d2c17bcbf7bae88f3749f412f7cd370e6e5d709f935e6bfef9053.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "da1a1eac4ccd57f14c51e06b9dd247b89358c6a19dc4f0488caa557dc3a462a1", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "5d64bb28c6d6dd3c401489c307e919f712775418ae952f0694f059d83fc36ef8", "generation_seconds_batch": 20.68309138098266, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2831003665924072, "target_z": -0.4589843484331378, "ce": 6.213237762451172, "genuineness": 4.457139015197754, "identity": 0.6170092225074768, "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_DRAM_reg/da1a1eac4ccd57f14c51e06b9dd247b89358c6a19dc4f0488caa557dc3a462a1.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": "daa0c1596c9cfd0cf85d8d2620dd8390f5b9a4ea18c1748658fb0d5186af6298", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "5694e1fdc260a586cb095365bf1aa5ac1c4567a619d6446c1a8c37918be67c7c", "generation_seconds_batch": 14.225342463003471, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.90083646774292, "target_z": 0.9609375245232574, "ce": 5.919358730316162, "genuineness": 4.124508857727051, "identity": 0.23256386816501617, "hypothesis": "Je bericht kwam nadat ik vertrokken was. Ik las het twee keer en legde mijn telefoon weg. Zijn 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_DRAM_reg/daa0c1596c9cfd0cf85d8d2620dd8390f5b9a4ea18c1748658fb0d5186af6298.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "06570aa5140201f70c20e8ad86a5f86db27204d5dcc00afffacbc0b033221db2", "generation_seconds_batch": 14.661946450010873, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9840450286865234, "target_z": 0.15624990190697036, "ce": 6.427689552307129, "genuineness": 4.143336772918701, "identity": 0.4297087788581848, "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_DRAM_reg/db192910e9ae67de251fb2a990940eab881f66f627c78b50b638ce61f5a63679.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "db92813fffbaff9accb69e06827da4dab6157dfad332e3688eb70d1d29b23e3b", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "b42fdb1d198933c65187c8f54542ca819437c5158bb050ee7e96b1246a01f34e", "generation_seconds_batch": 21.61246821191162, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.651611566543579, "target_z": 0.742187452588369, "ce": 6.054384231567383, "genuineness": 4.532451152801514, "identity": 0.3149947226047516, "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_DRAM_reg/db92813fffbaff9accb69e06827da4dab6157dfad332e3688eb70d1d29b23e3b.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": "dc6d9efed0a40703e9aa91e76713e1b17025e875b603378a8eb4cf995ac756e0", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Escribí el número en el sobre.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1c6cb92ee11c6f3881dbd8dd61827538a212ab0ee02b674018d57f533e993c3f", "generation_seconds_batch": 23.087157826987095, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.212367534637451, "target_z": 1.2343750621255853, "ce": 5.692992687225342, "genuineness": 4.093128204345703, "identity": 0.43953564763069153, "hypothesis": "Describí 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_DRAM_reg/dc6d9efed0a40703e9aa91e76713e1b17025e875b603378a8eb4cf995ac756e0.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "dc813bf453b0b67662612287f01a07a9bf4e1296f8859f27220d99fce6599681", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "59bc4264c01fa36f8b1b26ef903563cda7a18ffca996d10dd5913569df69980f", "generation_seconds_batch": 25.079145175986923, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.08775520324707, "target_z": 1.1250001307907063, "ce": 6.1894097328186035, "genuineness": 4.193545341491699, "identity": 0.45768529176712036, "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_DRAM_reg/dc813bf453b0b67662612287f01a07a9bf4e1296f8859f27220d99fce6599681.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "dd3f83aba8ecea27a194c33f15e63b5e7c8cc4fdbb3e7c1206e450a942e3b1b2", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "9b518ecb636e0e0eea4903cf524251dee39c7812f628e49712e94f60a1605e76", "generation_seconds_batch": 15.274529071059078, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8855791091918945, "target_z": 0.06982424031003047, "ce": 6.3641486167907715, "genuineness": 4.532451152801514, "identity": 0.40372979640960693, "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_DRAM_reg/dd3f83aba8ecea27a194c33f15e63b5e7c8cc4fdbb3e7c1206e450a942e3b1b2.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "dd6fc704afc09f6aba7dc1e9203364d9bb6a4a1ed2859037304df74f95db9125", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "6ac2a25392dca5e9cb60c0273d683426400a1aa88e412776954cab68de04c391", "generation_seconds_batch": 19.869309452013113, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.893923759460449, "target_z": 0.07714851985727281, "ce": 6.316492557525635, "genuineness": 4.268857479095459, "identity": 0.3132331669330597, "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_DRAM_reg/dd6fc704afc09f6aba7dc1e9203364d9bb6a4a1ed2859037304df74f95db9125.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "dda2af51cac8e43fb75066874b8b7d35cff69a970c55832d6fe476955b46545f", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "a9dc0552d12a01678a9e277e9ccf0d60d610699d76a310679ae6c434cfe880d0", "generation_seconds_batch": 20.73877645994071, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.710817337036133, "target_z": 1.6718752059953623, "ce": 5.534139156341553, "genuineness": 4.049196243286133, "identity": 0.09656273573637009, "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": 0.03999999999999915, "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_DRAM_reg/dda2af51cac8e43fb75066874b8b7d35cff69a970c55832d6fe476955b46545f.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "ddc2fb8102b22812f76c49f05c2db5cb51c625ab5c8d6f9aac81613dc0e64811", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "b471cca1a95be8e97105eb1040250f80e2695dcb1364df32b30689664e94001e", "generation_seconds_batch": 16.656995277036913, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.558152198791504, "target_z": 0.6601561494546446, "ce": 6.0146708488464355, "genuineness": 4.507347106933594, "identity": 0.6036949753761292, "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_DRAM_reg/ddc2fb8102b22812f76c49f05c2db5cb51c625ab5c8d6f9aac81613dc0e64811.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "de5c6f0890eeea8619662be8441417a9a05d627bb7fc48e7f6563d06c8ae4672", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "6a2eeb4c9cc33add1ebfa5e566a54a1c69b9fb090cda6d75bdc719f0c8d93f70", "generation_seconds_batch": 21.65030979795847, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.561253070831299, "target_z": -0.2148437197546492, "ce": 6.038498878479004, "genuineness": 4.231201171875, "identity": 0.746705174446106, "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_DRAM_reg/de5c6f0890eeea8619662be8441417a9a05d627bb7fc48e7f6563d06c8ae4672.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": "de72496badf15153930960bea0951cd3e115deb3d4cfb8a6017f93ab6b53b6d3", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "3dcdfd145b19dff35b99e7427ca0a086ea8fdb9537de19c744e631ffbda62c39", "generation_seconds_batch": 13.750861493987031, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.523898601531982, "target_z": 1.5078125997279135, "ce": 5.577824115753174, "genuineness": 4.300237655639648, "identity": 0.519783616065979, "hypothesis": "Ton message est arrivé après mon débat. Je l'ai relu, puis j'ai rangé mon téléphone. Dis-moi quand nous pourrons parler.", "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_DRAM_reg/de72496badf15153930960bea0951cd3e115deb3d4cfb8a6017f93ab6b53b6d3.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "df24e3346b9c3e60f682c2f3c62f098af7e9646218a842d9d2dec0f6d1e1e83a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "f693620bd1a43a6c4ecd07650d8202420e252278047b29588365f48ff0b6cf4d", "generation_seconds_batch": 20.125410704058595, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6872150897979736, "target_z": 0.7734374329697631, "ce": 6.205295085906982, "genuineness": 4.080576419830322, "identity": 0.34785744547843933, "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_DRAM_reg/df24e3346b9c3e60f682c2f3c62f098af7e9646218a842d9d2dec0f6d1e1e83a.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": "df3b4af69303462170b6e8ac9f861388669db6971444d77e6821bb5b2c084d30", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "2bcbb6e6647bc7c39f4f7d2a6f797062ff388f3dd16139e4deef90da7ed53cfb", "generation_seconds_batch": 15.471985918004066, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.434889793395996, "target_z": 1.4296876487744283, "ce": 5.673136234283447, "genuineness": 3.5439748764038086, "identity": 0.5735260248184204, "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_DRAM_reg/df3b4af69303462170b6e8ac9f861388669db6971444d77e6821bb5b2c084d30.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": "df84fa89a77cd89c5dd43400ed385c9fe2403c8ed7cef2ad80ef1e92004a2149", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "3c982c62f36951b0fd235feead1dd00b5938502c76de78e848471465eee2b596", "generation_seconds_batch": 15.75897237402387, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4290895462036133, "target_z": 0.5468750752046561, "ce": 5.959072113037109, "genuineness": 4.268857479095459, "identity": 0.47872161865234375, "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_DRAM_reg/df84fa89a77cd89c5dd43400ed385c9fe2403c8ed7cef2ad80ef1e92004a2149.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "e0339dfb679baa755d6d2ab1aad19ddc02d8b20329071e7f4904b6f70a68ac25", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "882ff860dc4fdd2f7ef220c332a50165d116b2a4e972b5144e5b4e0e3a8798a1", "generation_seconds_batch": 14.563995521049947, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.903380870819092, "target_z": 0.08544923050072752, "ce": 5.903473377227783, "genuineness": 4.18099308013916, "identity": 0.48328572511672974, "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_DRAM_reg/e0339dfb679baa755d6d2ab1aad19ddc02d8b20329071e7f4904b6f70a68ac25.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "e0abd7bd928974ba87d38a52db57bd1f46f30a866a44f1eecc0afa6efa1537dd", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "4ad1ded59507d981d4465bc32802643eb4729184db06356d8c86d300c9a7c655", "generation_seconds_batch": 21.78752606897615, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.487820625305176, "target_z": -0.2792969612401219, "ce": 5.855817794799805, "genuineness": 3.835810661315918, "identity": 0.5203625559806824, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/e0abd7bd928974ba87d38a52db57bd1f46f30a866a44f1eecc0afa6efa1537dd.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": "e0e991e4b589c98881634f719f02f443133cb8e4df482a93104079dadef0cb38", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "6a7b761c6b87b9f397ffeaf34ad6fd2988b1dad263f82369eb30a3564551cf9b", "generation_seconds_batch": 15.106861322070472, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8138155937194824, "target_z": 0.006835854734006232, "ce": 6.173524379730225, "genuineness": 4.6705241203308105, "identity": 0.5394835472106934, "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_DRAM_reg/e0e991e4b589c98881634f719f02f443133cb8e4df482a93104079dadef0cb38.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "e13b4e29829ae04b5bc3221b17c4808755bdf25f27978026ec662ccf1b7a2caf", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "4f50ac1161cc04dc6500116496a03f7851f53dfa69e0406af1a843bc29986927", "generation_seconds_batch": 15.955625968053937, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.021873950958252, "target_z": 0.189453163011049, "ce": 6.268836498260498, "genuineness": 4.796044826507568, "identity": 0.4446428418159485, "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_DRAM_reg/e13b4e29829ae04b5bc3221b17c4808755bdf25f27978026ec662ccf1b7a2caf.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": "e1813e8193c352c863693624ff871bd30c535259f10992b79c0f347c0e7baeec", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2adb4517dd03389682ddd4d30d31ba4f82dbba02c37767b8bca90c848020d9ec", "generation_seconds_batch": 21.890564540051855, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.5626027584075928, "target_z": 0.6640625016348838, "ce": 5.919358730316162, "genuineness": 4.137060642242432, "identity": 0.3938209116458893, "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_DRAM_reg/e1813e8193c352c863693624ff871bd30c535259f10992b79c0f347c0e7baeec.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "e1a4cf84741a784137d3dde557f26d5e946004253aa79dee7cdb596b2fef8fa5", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "4d75700642fbdb6756dc969e320d7e1bfd89c18d6880e66823b926af1c705736", "generation_seconds_batch": 20.6980390480021, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7990736961364746, "target_z": -0.006103426779281999, "ce": 6.4912309646606445, "genuineness": 4.632867813110352, "identity": 0.5694065093994141, "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_DRAM_reg/e1a4cf84741a784137d3dde557f26d5e946004253aa79dee7cdb596b2fef8fa5.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": "e27a3109148337503a1b1f826daf48bfd3e14f967c671e61672920329f0d92a9", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "16299e9fca9ad56453d185980ba34b04335465fff3cd83b4d352d74985a8c5df", "generation_seconds_batch": 21.033840715070255, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.731719493865967, "target_z": 0.8124999084465057, "ce": 5.959072113037109, "genuineness": 4.042920112609863, "identity": 0.5735712647438049, "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_DRAM_reg/e27a3109148337503a1b1f826daf48bfd3e14f967c671e61672920329f0d92a9.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "e2b54b7d1cef0b09a864eaf2f6ea9b661a7a9059671a6642c79a0b8261609100", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "c405389da0a8acd3ee463a4aef5d1d193c0db1db5a74c629527c59b722012142", "generation_seconds_batch": 20.504979660967365, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6981041431427, "target_z": -0.0947265815055245, "ce": 6.268836498260498, "genuineness": 4.482243061065674, "identity": 0.642075777053833, "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_DRAM_reg/e2b54b7d1cef0b09a864eaf2f6ea9b661a7a9059671a6642c79a0b8261609100.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "e3544eebf28b93b1acae84ae8dd2f6d4a73df1728fc90db4cec12c05ccc01ef3", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "77c097682412a9db23a9944dfd5576933a6857960850e723ec32df8286d7d959", "generation_seconds_batch": 21.732971908990294, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.625784397125244, "target_z": -0.1582031826296549, "ce": 6.221180438995361, "genuineness": 4.268857479095459, "identity": 0.44302186369895935, "hypothesis": "Ik heb een kleine doosje gevonden. Het stond naast het oude raam. We kunnen er 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_DRAM_reg/e3544eebf28b93b1acae84ae8dd2f6d4a73df1728fc90db4cec12c05ccc01ef3.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": "e449387f476164c60260f56cb2c4292e6b765db0b9ad109ea3c99ef647c640e1", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Your message arrived after I had left.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "e902ea73ac699bf6714897b58822d6ba616d62adf322c63a3794a380c3c5b7ec", "generation_seconds_batch": 15.548620968009345, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.08775520324707, "target_z": 1.1250001307907063, "ce": 6.070269584655762, "genuineness": 3.9173994064331055, "identity": 0.36126038432121277, "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_DRAM_reg/e449387f476164c60260f56cb2c4292e6b765db0b9ad109ea3c99ef647c640e1.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": "e45b6fa1307971fd9aa53c6cfa38eff282b8b236497a61a7ed14b2a62006a6ab", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "95233f2ce93d45037e368ab277207ab111991c17b8ebc885a8976283cf283470", "generation_seconds_batch": 22.070300573017448, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.247971057891846, "target_z": 1.2656250425069795, "ce": 5.25614595413208, "genuineness": 3.8169827461242676, "identity": 0.5615633130073547, "hypothesis": "De sleutels liggen op de keukentafel. Ik keek inderdaad en daarna, naast het raam, neem ze mee als je weggaat.", "wer": 0.13636363636363635, "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_DRAM_reg/e45b6fa1307971fd9aa53c6cfa38eff282b8b236497a61a7ed14b2a62006a6ab.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "e462de635705d27a59689575ca5f36791bff440469fa9a64b9fb20b0b0db266f", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "185a1708dfdf1777d8a4f710402859685f72335589615451827777c8acf1c47e", "generation_seconds_batch": 20.73278114804998, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7951793670654297, "target_z": -0.009521563411415036, "ce": 5.752562999725342, "genuineness": 4.344170093536377, "identity": 0.725894570350647, "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_DRAM_reg/e462de635705d27a59689575ca5f36791bff440469fa9a64b9fb20b0b0db266f.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "e472a10e154278aa9acdbcfddac3c4071781e0994352fb0c351ab95aa4eaaa7a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "1f78163faaa6f450795afba103a08480c05fe9f7cdde240e335909d898a5892a", "generation_seconds_batch": 15.992547197965905, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7406206130981445, "target_z": 0.8203126128069841, "ce": 5.621508598327637, "genuineness": 4.482243061065674, "identity": 0.6319710612297058, "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_DRAM_reg/e472a10e154278aa9acdbcfddac3c4071781e0994352fb0c351ab95aa4eaaa7a.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": "e4d03229198126be91124abfd06d4a02554af0a2ce7341b4888c63ddf581d0cd", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1e181354383d92b9823d24a5c2c1149f245e13a71d4cab50b5e963af890c244c", "generation_seconds_batch": 22.070300573017448, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.927539110183716, "target_z": 0.9843750098093029, "ce": 5.863759994506836, "genuineness": 4.293961524963379, "identity": 0.211425319314003, "hypothesis": "De sleutels liggen op de keukentafel. Ik kijk in de laad en daarna naast het raam. 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_DRAM_reg/e4d03229198126be91124abfd06d4a02554af0a2ce7341b4888c63ddf581d0cd.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": "e4e684bc94ea680383f19a81bcf429b49497bfcd7deb7a1200a99ceaafd8e0bd", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "1d5eb94743a88201fcf7e679c74c545a077deaceb3187ebdcb9f371c5475fe5e", "generation_seconds_batch": 14.85177109495271, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9898452758789062, "target_z": 1.0390624754767426, "ce": 5.748591423034668, "genuineness": 4.162164688110352, "identity": 0.7511498928070068, "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_DRAM_reg/e4e684bc94ea680383f19a81bcf429b49497bfcd7deb7a1200a99ceaafd8e0bd.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "e4f6d20841a2959a3b1fc643cb0a86abea2acfd1f7ef6b6b2531c72e061982e9", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "69f2cf81211a05923b8c7dbb736ca8cb99c41adfb2aee30dae7d9de48a13081c", "generation_seconds_batch": 16.067865638062358, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8385300636291504, "target_z": 0.9062498495906879, "ce": 5.919358730316162, "genuineness": 4.381826400756836, "identity": 0.69903963804245, "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_DRAM_reg/e4f6d20841a2959a3b1fc643cb0a86abea2acfd1f7ef6b6b2531c72e061982e9.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "e568e900c8d9e845bf78c3fbd35509fcd7436b3d7d7946a368135140106e0835", "bucket": "vn_S_DRAM_reg", "phase": "wording", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "0332928bf07f61a225693885042eb8d098f570441d7530de5d402e57f3d8d814", "generation_seconds_batch": 15.074661899008788, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5712666511535645, "target_z": -0.2060545842979584, "ce": 6.213237762451172, "genuineness": 4.281409740447998, "identity": 0.45504331588745117, "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_DRAM_reg/e568e900c8d9e845bf78c3fbd35509fcd7436b3d7d7946a368135140106e0835.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "e62536bc34d8a6b455428b37ff0473a5f03c152065c7085c32d18f3bf5a2b4fa", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "1d719709951ce08b3ff642432f0a6782b01f5c33d59986a34349251ad52eb967", "generation_seconds_batch": 20.84286193700973, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.577942132949829, "target_z": -0.20019536992529444, "ce": 6.221180438995361, "genuineness": 4.432034492492676, "identity": 0.563779354095459, "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_DRAM_reg/e62536bc34d8a6b455428b37ff0473a5f03c152065c7085c32d18f3bf5a2b4fa.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "e658ab028a60643a70d221589bc7816f18f80df1c97b9be8d91ef2417d2af154", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "c74b51340200f18da5153221aeadabea0fd4363fe10013e319bcc6dcbad86312", "generation_seconds_batch": 15.804306306992657, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2777743339538574, "target_z": 0.41406244931860137, "ce": 6.213237762451172, "genuineness": 4.507347106933594, "identity": 0.5526734590530396, "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_DRAM_reg/e658ab028a60643a70d221589bc7816f18f80df1c97b9be8d91ef2417d2af154.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "7a8b2dd0915b5df7d945ec47302a9491b063d181668402a6c601f96379fdc54f", "generation_seconds_batch": 15.135061737964861, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.289776086807251, "target_z": -0.45312492479534394, "ce": 6.356205940246582, "genuineness": 4.086852550506592, "identity": 0.3553027808666229, "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_DRAM_reg/e6f1e80cc6606ef00fdc322d5829b94fd65ad80a74ab015f88321e5dc194804c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "e7bea2042c757b73b5ab247f644c151e31cda1eec3f469364b03e6ff83a6b2c8", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "fd0d5e94f5c1f7c160a2a813e1c7cc3de154f7280ec85e4be9d17d650b61045b", "generation_seconds_batch": 21.03318190993741, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.3013763427734375, "target_z": 1.3125000130790707, "ce": 5.804190158843994, "genuineness": 4.394378662109375, "identity": 0.5504336953163147, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/e7bea2042c757b73b5ab247f644c151e31cda1eec3f469364b03e6ff83a6b2c8.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "e81bdae6d48caf62881f6ddfcfb729d543214ec3abe065a5fe164b92e54317e6", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "13269334c718ee3e5135960ec0d7427ccb49c2df6a0799cf8056954d5cf71763", "generation_seconds_batch": 21.534069605986588, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9528920650482178, "target_z": 0.1289062737058155, "ce": 6.292664527893066, "genuineness": 4.193545341491699, "identity": 0.400579571723938, "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_DRAM_reg/e81bdae6d48caf62881f6ddfcfb729d543214ec3abe065a5fe164b92e54317e6.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": "e84f0a30e694142f9218a7f685873a64d639f8b946b906a83106cd2755d73b55", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Escribí el número en el sobre.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "55801e0f8849a46f03777da9fbb31ff2e5cb64311f9f31eb5668380363e936ca", "generation_seconds_batch": 16.103637556079775, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.363682746887207, "target_z": 1.3671876880116403, "ce": 5.593709468841553, "genuineness": 3.9676074981689453, "identity": 0.337350070476532, "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_DRAM_reg/e84f0a30e694142f9218a7f685873a64d639f8b946b906a83106cd2755d73b55.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": "e86575080564fd93cb3fae367a693499d5a1e9b89b273f3b24e2a73193b38313", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "8a8ac897d4093b55ab8b47545840a01f394339bfd7b885fefa8c6d671230146a", "generation_seconds_batch": 15.934285057010129, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.408186912536621, "target_z": 1.406249954223253, "ce": 6.046441555023193, "genuineness": 4.13078498840332, "identity": 0.36865752935409546, "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_DRAM_reg/e86575080564fd93cb3fae367a693499d5a1e9b89b273f3b24e2a73193b38313.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "e96fb0b1796746bf6d64c848aeb68f9ccb2ff0fb03ad43fb9faa6c8005e5b844", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "4d73805fc1604f0bf915a27731013a8822608d9f6f8a8a0c09be3924cf8e70d8", "generation_seconds_batch": 20.445291862939484, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.708117723464966, "target_z": -0.08593744604883369, "ce": 6.165581703186035, "genuineness": 4.545003414154053, "identity": 0.8048818111419678, "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_DRAM_reg/e96fb0b1796746bf6d64c848aeb68f9ccb2ff0fb03ad43fb9faa6c8005e5b844.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "e9902099f96c95486dc383eeedee86d4fd8525f9a4342b9fa95654a8f7e73ad2", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "57af714352d6966f9158ef4d888af9a68add399a292d1acc604887351f92c61e", "generation_seconds_batch": 20.879176541930065, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.408186912536621, "target_z": 1.406249954223253, "ce": 5.244231700897217, "genuineness": 3.5690791606903076, "identity": 0.18362437188625336, "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_DRAM_reg/e9902099f96c95486dc383eeedee86d4fd8525f9a4342b9fa95654a8f7e73ad2.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": "e9a63fdb1d33972f20039163976b41047d42c26d019ffc99deb26a659c8e6fa0", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "bbad7350c5f9100cf7dc997d6349ac466d01cd7f8e36897616240ba35eca3df1", "generation_seconds_batch": 21.42373379005585, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9898452758789062, "target_z": 1.0390624754767426, "ce": 5.685050010681152, "genuineness": 4.281409740447998, "identity": 0.8009656667709351, "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_DRAM_reg/e9a63fdb1d33972f20039163976b41047d42c26d019ffc99deb26a659c8e6fa0.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "e9ba3317ec61625458d3754c0204a59b09d0e11ed9b7b42468244c2a910c8d6d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "cf8289c60f3437225aaa10785cc023b2f487ef09abf23f90bc9b625e950b24ca", "generation_seconds_batch": 22.26378329109866, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6658384799957275, "target_z": -0.12304685006802163, "ce": 6.1894097328186035, "genuineness": 4.7081804275512695, "identity": 0.7850601673126221, "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_DRAM_reg/e9ba3317ec61625458d3754c0204a59b09d0e11ed9b7b42468244c2a910c8d6d.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "88ca899d551fdfa8422c9d337a6b7371eb1ec711d3638455b1dc17625d859fc2", "generation_seconds_batch": 15.704136588959955, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.196316719055176, "target_z": -0.5351562279290684, "ce": 6.3641486167907715, "genuineness": 4.419482707977295, "identity": 0.4351227283477783, "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_DRAM_reg/e9d266c3907611a53d0c5f7688a7e231bba29ead6f87b8ae8d29cb37ab2fae41.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "ea2b754377c74a74dafd53ba084b0764d8b5ddb73bc23ba0fac30662c7927046", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "ce075732b31f98730100a4c4ed77812cd9f88968a41db85b7cf8e4907e29cba3", "generation_seconds_batch": 16.426489592064172, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.194565773010254, "target_z": 1.2187500719348885, "ce": 5.792276382446289, "genuineness": 4.231201171875, "identity": 0.6651053428649902, "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_DRAM_reg/ea2b754377c74a74dafd53ba084b0764d8b5ddb73bc23ba0fac30662c7927046.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "ea4e12567a4b253ebacfce0aa18a00956422bfaa3610e9cb07d103b7293c6004", "bucket": "vn_S_DRAM_reg", "phase": "wording", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "8d004aa16328c9f241cc5644bd07ece839a9beedfd7d2ca7d7ac1152c69857f8", "generation_seconds_batch": 14.487498109927401, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4010369777679443, "target_z": -0.35546884073605245, "ce": 6.443574905395508, "genuineness": 4.545003414154053, "identity": 0.40585047006607056, "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_DRAM_reg/ea4e12567a4b253ebacfce0aa18a00956422bfaa3610e9cb07d103b7293c6004.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "ea5f58f70dee3a31d1d47e319d5cd405476d98ba4513a4d9ebde09ecaae70545", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "54164c203d1e162cea4554b5c87483e242b7089bcf0c0853d264469e6b12255c", "generation_seconds_batch": 21.192304180935025, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.415738105773926, "target_z": 0.5351562279290684, "ce": 6.157639026641846, "genuineness": 4.2374773025512695, "identity": 0.21220313012599945, "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_DRAM_reg/ea5f58f70dee3a31d1d47e319d5cd405476d98ba4513a4d9ebde09ecaae70545.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "eb4ce7d62c260471daa54a21d6b091c5dc42d8ab01f030eb82c040e47bf57bb9", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "fc1431bbaf57d1accd39b1ac57251dd0abb34a24f3e7d276d7548233d71139df", "generation_seconds_batch": 21.65460466197692, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.043250560760498, "target_z": 1.0859374460488338, "ce": 5.418970584869385, "genuineness": 4.18099308013916, "identity": 0.3798374533653259, "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_DRAM_reg/eb4ce7d62c260471daa54a21d6b091c5dc42d8ab01f030eb82c040e47bf57bb9.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "eb621a9b7eb713ecf4b8b666b84e0f1c7690284a1a2c48f097d701452b33deae", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "9388dff1fbbece13e39e99a586f7659390eb5968b4a9b016210ab0469867c3f8", "generation_seconds_batch": 16.30342573300004, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8563320636749268, "target_z": 0.9218750490465148, "ce": 5.677107334136963, "genuineness": 4.369274139404297, "identity": 0.8309155702590942, "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_DRAM_reg/eb621a9b7eb713ecf4b8b666b84e0f1c7690284a1a2c48f097d701452b33deae.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 128, "seconds_generated": 10.24, "mp3_sha256": "c037637fdde93f5f76facba35ee5334396ba78cff3a2ea8ce6621ac2ffb8aa3b", "generation_seconds_batch": 15.261628977954388, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5623655319213867, "target_z": -0.21386728865843682, "ce": 6.117925643920898, "genuineness": 4.507347106933594, "identity": 0.7222753763198853, "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_DRAM_reg/ebe89981a89e9506d02bacdfa583f9bfe1f04faa699193749ab6419de85c588f.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "ebf94dc147eabecc4f79a095493c64e6d5fc537bee544b953a27e359370d7313", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "43e19d85e142bd5fedf0396a5a690ba35f408204fbc43c0fc14ca43355127096", "generation_seconds_batch": 19.5041069079889, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.505622386932373, "target_z": -0.26367197104942486, "ce": 6.24500846862793, "genuineness": 4.507347106933594, "identity": 0.6278509497642517, "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_DRAM_reg/ebf94dc147eabecc4f79a095493c64e6d5fc537bee544b953a27e359370d7313.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "ec0601913ff222399430393569967f7ca4eeb741e562edbf7196583d75cba3cf", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "5c92be60b305921ccaac21087b7be3e21ead9d8232eb55f0dfb881a9bd536730", "generation_seconds_batch": 15.397843602928333, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6160080432891846, "target_z": 0.7109374722069749, "ce": 6.149696350097656, "genuineness": 4.281409740447998, "identity": 0.4411490559577942, "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_DRAM_reg/ec0601913ff222399430393569967f7ca4eeb741e562edbf7196583d75cba3cf.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": "ec360d054d20704c0d75c16985fd9a0b035482cb29d877cea91715d41611d6f9", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "21ae3806aa7d8200ef87174348e658f2c4d86b8ade84ae5a2884a162f1ead92b", "generation_seconds_batch": 21.146918909973465, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.918638229370117, "target_z": 0.9765625147139545, "ce": 4.853055477142334, "genuineness": 3.9676074981689453, "identity": 0.39213797450065613, "hypothesis": "Las llaves están en la mesa de la cocina. Miren, 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_DRAM_reg/ec360d054d20704c0d75c16985fd9a0b035482cb29d877cea91715d41611d6f9.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "ec4cad57f80c65ba44a59cc7bcc7ddfa0a8c812e9cc39490c9c700706c503b4d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "477717bfb0d0aaca8633f956616af862c53b90120c91397d35e0126f3988e54b", "generation_seconds_batch": 21.771939640981145, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.073054075241089, "target_z": 0.23437506212558545, "ce": 6.284721851348877, "genuineness": 4.331617832183838, "identity": 0.7160360217094421, "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_DRAM_reg/ec4cad57f80c65ba44a59cc7bcc7ddfa0a8c812e9cc39490c9c700706c503b4d.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": "ec523046c5fd51b4fc6a13631708ab03045675566485851f10b5b49ccdfb1484", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The keys are on the kitchen table.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9922915a2abef3d7ab448e99c2ccd1dd091d5db723f4d0991b4e5a5bb189089c", "generation_seconds_batch": 22.325334085035138, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7673230171203613, "target_z": 0.8437498888278997, "ce": 6.006728172302246, "genuineness": 4.300237655639648, "identity": 0.3818356692790985, "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_DRAM_reg/ec523046c5fd51b4fc6a13631708ab03045675566485851f10b5b49ccdfb1484.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "ec7d9ed0390bd3a7bd93f2c4bada504950d1e25ae979f6591830e65a8a3d98c2", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "35480004815e097f89401713177429defdb1c041867bf81d08f26c2d31b81dd6", "generation_seconds_batch": 24.770493646035902, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.575953960418701, "target_z": 0.6757811396453416, "ce": 5.863759994506836, "genuineness": 4.557555198669434, "identity": 0.7150033116340637, "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_DRAM_reg/ec7d9ed0390bd3a7bd93f2c4bada504950d1e25ae979f6591830e65a8a3d98c2.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "ed39e68321d683db5c32e84ef93603204e15b2aebf3748b43e0f9bc9c13806fd", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "8ddca2e09b403965d997728ac68ba9dcc9d73159dfce2663bd28a4ac7915e9c2", "generation_seconds_batch": 21.34575086995028, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.384584903717041, "target_z": 0.5078123904627836, "ce": 5.855817794799805, "genuineness": 4.432034492492676, "identity": 0.7437825202941895, "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_DRAM_reg/ed39e68321d683db5c32e84ef93603204e15b2aebf3748b43e0f9bc9c13806fd.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "ed64096c956b0b823c7e41406885be1036be89bf436485a51cda3041759eb506", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "0f200e0f338d96cf46b3e75fb89f716bfd0f18b5547119b34c139f7ffdbc6f53", "generation_seconds_batch": 22.19174836599268, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.4913957118988037, "target_z": 0.6015625408720957, "ce": 6.173524379730225, "genuineness": 4.400654315948486, "identity": 0.46312910318374634, "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_DRAM_reg/ed64096c956b0b823c7e41406885be1036be89bf436485a51cda3041759eb506.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "edee6d1c8965aceaf3d1c5718b0bc64c5d7bbb772855a0cd2ab259156dc3a223", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "72af50f73d9041487625546363859a2444c28d53b09aa015a376d485dbd67dce", "generation_seconds_batch": 15.640359251061454, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1042070388793945, "target_z": 0.2617186903267403, "ce": 6.1894097328186035, "genuineness": 4.362998008728027, "identity": 0.39294785261154175, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/edee6d1c8965aceaf3d1c5718b0bc64c5d7bbb772855a0cd2ab259156dc3a223.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "ee6b346deb70507ad911300623eecd56594581c726692b4701a0a295bab85416", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "b02a3721c27dc92ee7b19ed1ccc466d68acf6f62f3b855bf192ba380be24ce4e", "generation_seconds_batch": 21.071550339926034, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.212367534637451, "target_z": 1.2343750621255853, "ce": 5.1052350997924805, "genuineness": 4.093128204345703, "identity": 0.29910314083099365, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/ee6b346deb70507ad911300623eecd56594581c726692b4701a0a295bab85416.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "ee8cddf1c2dddc11c67348b55156c96ee17b1ebc2d10b2c4c67ea2ca298a8ef9", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "68a9c899dc04c0d465c946584faca6fb97b52b2430e0d07ceddae08d6ac7f72a", "generation_seconds_batch": 21.534069605986588, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.286675214767456, "target_z": 0.42187494441394985, "ce": 6.372091293334961, "genuineness": 4.457139015197754, "identity": 0.4432588219642639, "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_DRAM_reg/ee8cddf1c2dddc11c67348b55156c96ee17b1ebc2d10b2c4c67ea2ca298a8ef9.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": "eeea879fe5bda8f32a26adf2b6909cf34195f2def62825af4200df5b1cb0a3df", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The keys are on the kitchen table.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "3608ece4d1636922cb51ed5526ad458aecee409fd10c4d6505da3ab802f0c9d0", "generation_seconds_batch": 14.904511532979086, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.123358726501465, "target_z": 1.1562501111721002, "ce": 5.7684478759765625, "genuineness": 3.9927117824554443, "identity": 0.3869290351867676, "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_DRAM_reg/eeea879fe5bda8f32a26adf2b6909cf34195f2def62825af4200df5b1cb0a3df.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": "f1a0a9c31cf443dab27cfaa3f3f527b6abdd911297bd082237f2f3f288195d3d", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d293dfbc01a9118239d0529e0701e7a8795ab2cbd278c10812ead80cb5e3814c", "generation_seconds_batch": 22.110864380956627, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.265772819519043, "target_z": 1.2812500326976766, "ce": 5.474569320678711, "genuineness": 4.256305694580078, "identity": 0.6512222290039062, "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_DRAM_reg/f1a0a9c31cf443dab27cfaa3f3f527b6abdd911297bd082237f2f3f288195d3d.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "463df31531860489cee627b5390b0d42fef02aaa1aa57fb2729c86807130e6f4", "generation_seconds_batch": 14.92464466392994, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5734918117523193, "target_z": -0.20410151284040373, "ce": 6.300607204437256, "genuineness": 4.595211505889893, "identity": 0.7799628376960754, "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_DRAM_reg/f21bc0adf2df6aeda4aa5bc68657c60b57559ffed3b647118e970ebfd0ba6ad8.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "f22e83d152c412758135e4c17319ed3d1b767a7cce56349e9e8649a5545da28f", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "23f2a00f0f9d8b8dae3b4ab2fae789cacc0b861df49a2b0bc08c1d9e928591e2", "generation_seconds_batch": 21.37001985195093, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7796030044555664, "target_z": -0.0231932728794275, "ce": 6.062326908111572, "genuineness": 4.482243061065674, "identity": 0.7566714882850647, "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_DRAM_reg/f22e83d152c412758135e4c17319ed3d1b767a7cce56349e9e8649a5545da28f.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "f24b71b914e04ec8f76a7a1da2dc451202a89f7101e8b14175a743145eabed17", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "1d83114ef7a84e75680b0ae07637f717cee3adf129806aaa34fe7db8f418e8b3", "generation_seconds_batch": 21.553491873899475, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.758422374725342, "target_z": 0.8359376029976812, "ce": 5.252174377441406, "genuineness": 4.425758361816406, "identity": 0.47176069021224976, "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_DRAM_reg/f24b71b914e04ec8f76a7a1da2dc451202a89f7101e8b14175a743145eabed17.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "f29fc8d9eb206d41838e22e02bc63e123634b4549bec1377df39b7efdf429df9", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "626ab1e0ef3df4c09a8d7ab8fc996dc802a2d4ab211f508d7707b29f3421ff55", "generation_seconds_batch": 20.895830499008298, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.478044271469116, "target_z": 0.5898436935965079, "ce": 6.157639026641846, "genuineness": 4.419482707977295, "identity": 0.6972692608833313, "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": 0.03999999999999915, "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_DRAM_reg/f29fc8d9eb206d41838e22e02bc63e123634b4549bec1377df39b7efdf429df9.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "f2a75a396846da0a493cf9185dbd0094fdffd45c623d97ade29f8f9f99592569", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "bac1e44320db06860bea55ddedf7d61800e3586abbf3e7dec539e168a69e7335", "generation_seconds_batch": 21.882755789905787, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.796013832092285, "target_z": -0.008789135456690803, "ce": 6.23706579208374, "genuineness": 4.570107460021973, "identity": 0.7216150760650635, "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_DRAM_reg/f2a75a396846da0a493cf9185dbd0094fdffd45c623d97ade29f8f9f99592569.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "f2e5612953f1ea7a23fb55aa62640da6341657687781b524b048d44e55231251", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "6ffabb40c4e95191887285b3cf2f807d6377c04ff4dc885af26daec30cfb3367", "generation_seconds_batch": 20.55909958994016, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8118276596069336, "target_z": 0.8828125735697723, "ce": 5.95112943649292, "genuineness": 4.306513786315918, "identity": 0.3375769257545471, "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_DRAM_reg/f2e5612953f1ea7a23fb55aa62640da6341657687781b524b048d44e55231251.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "f3b01f6ec45ec8cef1f8766475acb0bef0e071c6fb3ebf94f63cc6cac00ae326", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "8952b8b1eb469069127544d18285fc18c199116d1793e1817cd60ea319f4fbab", "generation_seconds_batch": 16.032449984923005, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.629122257232666, "target_z": -0.15527347081075799, "ce": 6.252951145172119, "genuineness": 4.419482707977295, "identity": 0.7681310176849365, "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_DRAM_reg/f3b01f6ec45ec8cef1f8766475acb0bef0e071c6fb3ebf94f63cc6cac00ae326.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "f42237805fc44b75cd30fb0399c4556ff2522ad255eb761d1947cc46994c82ef", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "336e1a95a6a0d5398df1f8a72bccedbd95324e68f4cf6c241f53fbecae286d44", "generation_seconds_batch": 21.54336655093357, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.142035961151123, "target_z": 0.29492195143081895, "ce": 6.260893821716309, "genuineness": 4.381826400756836, "identity": 0.5051374435424805, "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_DRAM_reg/f42237805fc44b75cd30fb0399c4556ff2522ad255eb761d1947cc46994c82ef.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "f42f6d5f41023b71d1ffa37a8636aeb1631d8971d9eb0879ebc81be2dfd5523e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "7358f3d0376821e215bf4d9b87df7bf7d764466cd72c76d4b923560e47f7d4c1", "generation_seconds_batch": 15.768428527982906, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.166513442993164, "target_z": 0.3164063652593098, "ce": 6.173524379730225, "genuineness": 4.243753433227539, "identity": 0.7879396080970764, "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_DRAM_reg/f42f6d5f41023b71d1ffa37a8636aeb1631d8971d9eb0879ebc81be2dfd5523e.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "f47b64f8dbc63330431aedb33bfaffed111427c585f4bd09550fdd717e666813", "bucket": "vn_S_DRAM_reg", "phase": "wording", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "c10e58e9ae8583b17241f121f1d02bfa8b23df76a7888f9e5995bd518c1a8128", "generation_seconds_batch": 16.879051913972944, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1607131958007812, "target_z": -0.5664062083104624, "ce": 6.117925643920898, "genuineness": 4.582659721374512, "identity": 0.463248610496521, "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_DRAM_reg/f47b64f8dbc63330431aedb33bfaffed111427c585f4bd09550fdd717e666813.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "f4ca2e15f7e48afc8f818e880a4da32ed670ecf8fba2fc0222c079a3a4aed25e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "a76cb5e465cf9291c5a5db6c0cb4d3ed9d95920b5e4d46adfa959169c15cf47f", "generation_seconds_batch": 15.6884396529058, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.357882499694824, "target_z": 0.4843751144418679, "ce": 6.340320587158203, "genuineness": 4.319066047668457, "identity": 0.3558076024055481, "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_DRAM_reg/f4ca2e15f7e48afc8f818e880a4da32ed670ecf8fba2fc0222c079a3a4aed25e.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "f4d3fad4f6e5995bac7269f8a4b1d3a94ae50b3f4c634ea9890ef81c9a7c78d7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "2bf2a79d792a9e7e631ca6bafa90604f2946d0c262f2e4eb0304d0d7d229ae39", "generation_seconds_batch": 21.117827340960503, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.319178104400635, "target_z": 1.3281250032697676, "ce": 6.165581703186035, "genuineness": 4.250029563903809, "identity": 0.6508154273033142, "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": 0.03999999999999915, "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_DRAM_reg/f4d3fad4f6e5995bac7269f8a4b1d3a94ae50b3f4c634ea9890ef81c9a7c78d7.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-1", "key": "f558d16c88307153237bc037a1d57774f98c22c755e735cd40d4c7c06a341c66", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "2d008df8ee6ad6cb91ffeb39e33aa33eef55937ee2696dd160f97d260d8daf00", "generation_seconds_batch": 20.84286193700973, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.534550428390503, "target_z": -0.23828120504069475, "ce": 6.3641486167907715, "genuineness": 4.394378662109375, "identity": 0.5094165802001953, "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_DRAM_reg/f558d16c88307153237bc037a1d57774f98c22c755e735cd40d4c7c06a341c66.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "f5e9c19aeb0f67b9f2fc738602e46ad0d0350feebf4b334694e194cbbfe39664", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "dc7c1bd9212af54f6c5d5b8908d87678ca6748ce996b3dd9f975604c1b57f401", "generation_seconds_batch": 22.730503169004805, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.675213813781738, "target_z": 1.6406252256139682, "ce": 5.657250881195068, "genuineness": 3.898571252822876, "identity": 0.40138956904411316, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude graam. We kunnen er samen in kijken.", "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_DRAM_reg/f5e9c19aeb0f67b9f2fc738602e46ad0d0350feebf4b334694e194cbbfe39664.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "f6401882171421be23d3ff8c3ab995af0f8993535ce2d04f3ee7d38dcce1606a", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "27448385e2c38a946dc5237a24abb9b5398dc4b82e7ec49557a74efc7ecc35b4", "generation_seconds_batch": 20.811232559964992, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.432190179824829, "target_z": -0.3281250032697677, "ce": 6.046441555023193, "genuineness": 4.375550270080566, "identity": 0.7591218948364258, "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_DRAM_reg/f6401882171421be23d3ff8c3ab995af0f8993535ce2d04f3ee7d38dcce1606a.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-1", "key": "f6b631c1f462aead4f77e2791b1cabbc84cbb691475e2ddac9fbc32cfb1d715d", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] He encontrado la cajita.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "4697a19ce846cc8d218c4c5f2ca6e6e1547b650d133ea15aa7258dd05d96e884", "generation_seconds_batch": 15.839338808087632, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.630234956741333, "target_z": -0.1542968304494157, "ce": 6.23706579208374, "genuineness": 4.494794845581055, "identity": 0.6983460187911987, "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_DRAM_reg/f6b631c1f462aead4f77e2791b1cabbc84cbb691475e2ddac9fbc32cfb1d715d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "f6c3c0d4d6c94b3592108b3679b57347676257582d0211902d4dae40514447c5", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "9d20b9b87736ab73e9816037e44d61cd7e6f6ab63b86ca260e98c5c6c10b8a5c", "generation_seconds_batch": 20.887976292986423, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.069953441619873, "target_z": 1.1093751406000092, "ce": 5.3792572021484375, "genuineness": 4.243753433227539, "identity": 0.4850371479988098, "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_DRAM_reg/f6c3c0d4d6c94b3592108b3679b57347676257582d0211902d4dae40514447c5.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "f6cb878c710c867985efd2c2081a3d54211b7d6a1fa2d49ff303b2643c1458f9", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "0932defbc037a82f89aaafbd6a967b11fd414c882ef69d3850a44b24b8947fe0", "generation_seconds_batch": 21.62540798506234, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2955760955810547, "target_z": 0.4296874395092984, "ce": 6.062326908111572, "genuineness": 4.18099308013916, "identity": 0.3486839234828949, "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_DRAM_reg/f6cb878c710c867985efd2c2081a3d54211b7d6a1fa2d49ff303b2643c1458f9.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "f709c06a8b10bb0c54b085c5eb10a4864c17455fa71163f739c68b3875eb92ac", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "28bcf48c1e2e633abc6970717326f92e88a4b5f1794fb76a2d7891ae30927c2e", "generation_seconds_batch": 20.608124705962837, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.847431182861328, "target_z": 0.9140625539511663, "ce": 6.24500846862793, "genuineness": 3.948779582977295, "identity": 0.319633424282074, "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_DRAM_reg/f709c06a8b10bb0c54b085c5eb10a4864c17455fa71163f739c68b3875eb92ac.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 128, "seconds_generated": 10.24, "mp3_sha256": "a57c1bc774eae8d631c57255f2fe8eaba90f101bf66c50ad0a4c2bfab87d02d7", "generation_seconds_batch": 15.315051507088356, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.779881000518799, "target_z": -0.02294926973793937, "ce": 6.268836498260498, "genuineness": 4.494794845581055, "identity": 0.6970666646957397, "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_DRAM_reg/f7a65885d0e107603bdce9d92b5fc3b68a68524db7b0b06f79b6538794e8bf5c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "f81e88cd693d4727cc73c566e1257c02a56c17f83c2e977f2d52baff29551930", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "ebdb467099ae71a0ed8722ade03d95b96ae861f9899f0196e6524d6ad3238e17", "generation_seconds_batch": 18.78322666697204, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8919355869293213, "target_z": 0.9531250294279089, "ce": 5.327630043029785, "genuineness": 4.15588903427124, "identity": 0.4769861698150635, "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_DRAM_reg/f81e88cd693d4727cc73c566e1257c02a56c17f83c2e977f2d52baff29551930.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "f82e2d2f9e03dd4e22b1af01dacee7fe4d26e3a6400692630191c1a492aeb044", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "08ec7922d9a1bcf6a903640601c73c7871f0c1cc14a9b5dba9893f7e559f7740", "generation_seconds_batch": 22.321696840925142, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.8118276596069336, "target_z": 0.8828125735697723, "ce": 6.070269584655762, "genuineness": 4.168440818786621, "identity": 0.5460376143455505, "hypothesis": "Ik heb het kleine toosje gevonden. Het stond naast het oude raam. We kunnen er 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_DRAM_reg/f82e2d2f9e03dd4e22b1af01dacee7fe4d26e3a6400692630191c1a492aeb044.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_DRAM_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": "15e50efa96ed18f888cbd380780501495b12e9b524a55038ca5a0cc0d0b67177", "generation_seconds_batch": 15.670980485039763, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6845157146453857, "target_z": -0.9843750098093029, "ce": 6.356205940246582, "genuineness": 4.369274139404297, "identity": 0.6201926469802856, "hypothesis": "Die Straße führt hinter der Brücke nach links. Geh weiter bis zum kleinen Holstor. Ich warte auf der anderen Seite.", "wer": 0.14285714285714285, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_DRAM_reg/f86feddc7af334e5c3d6691e715ece3168f8e7618bd3b63ca7d0171f412fdf9d.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "f87f82874ed9972380d897425f9e22617782fe089e7a5c3a189b80d00d20045b", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "da40af39099bff34d167afeee02a1c0e7631f98f86f776822e6a201e336ca28e", "generation_seconds_batch": 19.73885453992989, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.190990686416626, "target_z": 0.3378905698226708, "ce": 5.9431867599487305, "genuineness": 4.65797233581543, "identity": 0.7090804576873779, "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_DRAM_reg/f87f82874ed9972380d897425f9e22617782fe089e7a5c3a189b80d00d20045b.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "f88135f16226c75849a3293fee64a7d95ea0f0e8b440c49f2745bf4588d7eabc", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "49c93cd3b175af217309bf13958dcc98e26dda125a1bd6110d8243afe48ddfbb", "generation_seconds_batch": 20.865469941985793, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.438865900039673, "target_z": -0.3222655796319738, "ce": 6.356205940246582, "genuineness": 4.419482707977295, "identity": 0.45657408237457275, "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_DRAM_reg/f88135f16226c75849a3293fee64a7d95ea0f0e8b440c49f2745bf4588d7eabc.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "f8acfb0dc764a20bd2f3e0c9b1cef4329b41cfbba3ea9fa76a4afd92ae06f433", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "1ffffb8429a9cfa5e93ff4e16b68c61e12a70525bc70a7b3ba1b2cedf032496f", "generation_seconds_batch": 20.564424208016135, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.7851247787475586, "target_z": 0.8593748790185968, "ce": 5.808161735534668, "genuineness": 4.086852550506592, "identity": 0.3466942310333252, "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_DRAM_reg/f8acfb0dc764a20bd2f3e0c9b1cef4329b41cfbba3ea9fa76a4afd92ae06f433.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "f91556fe57a7d67094cdc1c9ff15f988b902107f7cd712e37d4b70f0aee48845", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "72d1ee0287bc0d70bd3111ec9f286329205d175abc07e0d38e62d070dfda6f42", "generation_seconds_batch": 21.304134557954967, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.114457607269287, "target_z": 1.1484374068116219, "ce": 5.5817952156066895, "genuineness": 4.055472373962402, "identity": 0.2980297803878784, "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_DRAM_reg/f91556fe57a7d67094cdc1c9ff15f988b902107f7cd712e37d4b70f0aee48845.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": "f98dc05eda76ac04e4cfedef78366b9ed0098c7cc9f1476a3b3c86e1ae4c8bd9", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "544fa4a45f364e0af37aaa2a65dc7738273f120d5e632101b2b9b8f33342f2da", "generation_seconds_batch": 15.106861322070472, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6369106769561768, "target_z": -0.14843740681162182, "ce": 6.181467056274414, "genuineness": 4.607763767242432, "identity": 0.5762720108032227, "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_DRAM_reg/f98dc05eda76ac04e4cfedef78366b9ed0098c7cc9f1476a3b3c86e1ae4c8bd9.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "fa6d0f2ad55bd5d53a5383cb64c6f9cbd26ed88cdf8b567075aa94a9c5f15b40", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "9cfd4217abe6eed8e77474847c60db66ceb8733628b7ce0668aad01c511afeb0", "generation_seconds_batch": 20.696722930995747, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3556571006774902, "target_z": 0.48242183371918335, "ce": 5.621508598327637, "genuineness": 4.381826400756836, "identity": 0.8243634700775146, "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_DRAM_reg/fa6d0f2ad55bd5d53a5383cb64c6f9cbd26ed88cdf8b567075aa94a9c5f15b40.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "faa069c90de45f574f9ccc27c6dffa5520e8a062d9cec83fb7421ed8a1ab1571", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "b777967e8ed97f9caf941d27579e8dde4d20af720872b24f4b20e49bc86cf056", "generation_seconds_batch": 21.56047468702309, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.429965019226074, "target_z": -0.3300780747273223, "ce": 6.141753673553467, "genuineness": 4.281409740447998, "identity": 0.7866479754447937, "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_DRAM_reg/faa069c90de45f574f9ccc27c6dffa5520e8a062d9cec83fb7421ed8a1ab1571.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "fab31cd7c8c97c36a5c8e4e18871c00d49367a02596ad6dc21138dfd0eca0c81", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "abbecf87fe26f99e0c5f9b224d3a23233e7b11174031c6bc96fe127af446fc99", "generation_seconds_batch": 14.958771220990457, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.923088550567627, "target_z": 0.9804686576290638, "ce": 6.0305562019348145, "genuineness": 4.369274139404297, "identity": 0.6869776248931885, "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_DRAM_reg/fab31cd7c8c97c36a5c8e4e18871c00d49367a02596ad6dc21138dfd0eca0c81.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "fb3eaba7be9af4ee5c6a6f14b85ed3226c56f1ce64520c5068c4358488dc68bb", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "a36e65fcf8b315c9c31cc6d83354febdc65c9ef6de39dbbde0fea089440fd135", "generation_seconds_batch": 21.20461635896936, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9809443950653076, "target_z": 1.0312499803813941, "ce": 5.289902210235596, "genuineness": 4.243753433227539, "identity": 0.14966337382793427, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/fb3eaba7be9af4ee5c6a6f14b85ed3226c56f1ce64520c5068c4358488dc68bb.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": "fb466d75e6038cc30de98b7d02109b6f8036d27e716beb765ae0aad1e0e0a468", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I wrote the number on the envelope.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "62c610dd494a117ef7bee7ee26b18d0ac348c275dc47e5520460e4081ad8726c", "generation_seconds_batch": 15.44706453208346, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.425988674163818, "target_z": 1.4218749444139498, "ce": 6.260893821716309, "genuineness": 3.7793264389038086, "identity": 0.3618370294570923, "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_DRAM_reg/fb466d75e6038cc30de98b7d02109b6f8036d27e716beb765ae0aad1e0e0a468.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-0", "key": "fb6ddfd3c8f9a6ac769a61a28067452f784fdd1c4a907f471100b6f091fa2f28", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] The door is still open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] I put your coat on the chair.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "ce3fadd059eb50f28b33cc052d875dd3d0c5d0fb47e066ad6c424dcd0e19f60f", "generation_seconds_batch": 20.504235600004904, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9631426334381104, "target_z": 1.015624990190697, "ce": 6.173524379730225, "genuineness": 4.174716949462891, "identity": 0.44469624757766724, "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/manual-comparison-v1/audio/vn_S_DRAM_reg/fb6ddfd3c8f9a6ac769a61a28067452f784fdd1c4a907f471100b6f091fa2f28.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-0", "key": "fb8e034dee0dee00f0058fb81cb44cb20a2212db99d89b48ae829ec9f931b8b7", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La porte est encore ouverte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "c91450fc5763202fa14b92ba460d62cabe0cdd88df9790f0e45a3680ac76e524", "generation_seconds_batch": 19.80487561505288, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.781132698059082, "target_z": -0.02185062780585302, "ce": 5.990842819213867, "genuineness": 3.9299514293670654, "identity": 0.6465858221054077, "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_DRAM_reg/fb8e034dee0dee00f0058fb81cb44cb20a2212db99d89b48ae829ec9f931b8b7.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "fbd79dc5207eb26c43dfebad135f27317b0381a28424e90cb7ff2c79f5d7be02", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "2c4fa53df885ca665950a31cdd8491b2a098735277dd3085eac73ad8d3e2a460", "generation_seconds_batch": 24.32487613603007, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6942100524902344, "target_z": -0.09814450887252761, "ce": 6.229123115539551, "genuineness": 4.6203155517578125, "identity": 0.4459298849105835, "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_DRAM_reg/fbd79dc5207eb26c43dfebad135f27317b0381a28424e90cb7ff2c79f5d7be02.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "fc1b5116c1893c196465f49cd286a200b659ab1391a0e8f6bd04c15d66e86784", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "f5c8a9400ed990f77b5605a996c7059265ba604259b282e48c572eeeec00d405", "generation_seconds_batch": 20.22721594106406, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2822248935699463, "target_z": 0.4179688014988406, "ce": 6.403861999511719, "genuineness": 4.607763767242432, "identity": 0.3254551887512207, "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_DRAM_reg/fc1b5116c1893c196465f49cd286a200b659ab1391a0e8f6bd04c15d66e86784.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "fc26089748f0ec423464a59b584ca4c7b6ddff2bbb091b141cf6b253d4aca971", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "e53d636476c88bf83b287f9f8c64cde6cd7d08d9eebbb6b94204ad397c42ccf5", "generation_seconds_batch": 21.718066281056963, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.9809443950653076, "target_z": 1.0312499803813941, "ce": 5.411027908325195, "genuineness": 4.024092197418213, "identity": 0.354483425617218, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/fc26089748f0ec423464a59b584ca4c7b6ddff2bbb091b141cf6b253d4aca971.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "fc2bc77711c79a7460fd548f67e4e5f128a70cfefa4da78a2a5f9d13a1335666", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "d0b85da013e7fb7bee611b6695fc1f6d3a84a56f856b2e76b9c282365338ac87", "generation_seconds_batch": 22.68269412498921, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.286675214767456, "target_z": 0.42187494441394985, "ce": 5.95112943649292, "genuineness": 4.6203155517578125, "identity": 0.4478982985019684, "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_DRAM_reg/fc2bc77711c79a7460fd548f67e4e5f128a70cfefa4da78a2a5f9d13a1335666.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": "fc30ec2c7e7efb357bba1b00e053e3d863bff76f6adb93e43c80d58e8897a7d2", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9a409b23d2fbd310a487a588acfb62e6f691c728c809a37e36a91f5016e99370", "generation_seconds_batch": 22.243346164003015, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.167862892150879, "target_z": 1.1953123773837129, "ce": 5.935244083404541, "genuineness": 4.093128204345703, "identity": 0.6224322319030762, "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_DRAM_reg/fc30ec2c7e7efb357bba1b00e053e3d863bff76f6adb93e43c80d58e8897a7d2.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-de-0", "key": "fc3d59e9f8969f491326ca73892c6d584f734f8ca53b3f21d4cdb8060c38eda2", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Die Tür ist noch offen.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "391409894d68f5e9e5f5f2d95bc48551923b4f97e562a56875db2e5e2ecf4485", "generation_seconds_batch": 15.837719884933904, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.3534319400787354, "target_z": 0.4804687622616287, "ce": 6.23706579208374, "genuineness": 4.821148872375488, "identity": 0.5417646169662476, "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_DRAM_reg/fc3d59e9f8969f491326ca73892c6d584f734f8ca53b3f21d4cdb8060c38eda2.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-fr-1", "key": "fc436b41e4071472048296d0b351472c0fe9516f3c2977b365eaaacab1439a52", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "e0e8cee5a17a463f661572bc622ca83ca8820fa7fb3ce4451adf2b42448fa6fc", "generation_seconds_batch": 20.696722930995747, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.47939395904541, "target_z": 1.468749914986041, "ce": 5.331601142883301, "genuineness": 3.973883628845215, "identity": 0.7575921416282654, "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_DRAM_reg/fc436b41e4071472048296d0b351472c0fe9516f3c2977b365eaaacab1439a52.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-1", "key": "fca74f7ad22771361125ea8e3c9210ee56f6b6850ea1aab882e9f9e93d4ee21e", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Het stond naast het oude raam.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "9794395f1efebddb6cb607a1fb7e8813dee8d65f0e32013cec4dbc415910553e", "generation_seconds_batch": 22.321696840925142, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.602656841278076, "target_z": 0.6992188341965171, "ce": 6.109982967376709, "genuineness": 4.444586753845215, "identity": 0.48726341128349304, "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_DRAM_reg/fca74f7ad22771361125ea8e3c9210ee56f6b6850ea1aab882e9f9e93d4ee21e.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "fcab6ee8696760f2d4cd4b833ba7c2e6326db36e82f3b73bda5c2da1d146c791", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "93e4de76190e1af0c63b6e29bc86a14c096dea970a167bfface0201c7e9f6c70", "generation_seconds_batch": 20.0129872179823, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.6872150897979736, "target_z": 0.7734374329697631, "ce": 5.164804935455322, "genuineness": 4.20609712600708, "identity": 0.5034189820289612, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/fcab6ee8696760f2d4cd4b833ba7c2e6326db36e82f3b73bda5c2da1d146c791.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "fcfd24d9d155a3232073fe7b147e753cd139b858ad456f755dcc69e8613e587c", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "93660fc30c0db1ad06e13399191d585741018d633b787911bec09b0a3d62d3a5", "generation_seconds_batch": 23.309548681019805, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7865567207336426, "target_z": -0.017089846100145502, "ce": 6.09409761428833, "genuineness": 4.469690799713135, "identity": 0.4945084750652313, "hypothesis": "La puerta sigue abierta. Eh. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.06666666666666667, "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_DRAM_reg/fcfd24d9d155a3232073fe7b147e753cd139b858ad456f755dcc69e8613e587c.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "feba9be7fdca06302a0837a94f47a29c381dd35bfaea0985d399f2e4c5f36f34", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "d32ba47a86d2e9f2d0c1b13dd8603f30062f3384320c88685fbc749abf250081", "generation_seconds_batch": 20.0129872179823, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.47939395904541, "target_z": 1.468749914986041, "ce": 5.458683967590332, "genuineness": 3.8295347690582275, "identity": 0.294201135635376, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/feba9be7fdca06302a0837a94f47a29c381dd35bfaea0985d399f2e4c5f36f34.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-en-1", "key": "febb0bc91289e03206e205b129b8a6eacae493cac73c19db35e74946a6db5816", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] I found the little box.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It was beside the old window.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "fe55c3bbd4bdc32d305db3b647f94eafd6dca500de937d531531aa0924fb960b", "generation_seconds_batch": 20.865469941985793, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.752065896987915, "target_z": -0.047363186120197284, "ce": 6.356205940246582, "genuineness": 4.419482707977295, "identity": 0.4270322918891907, "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_DRAM_reg/febb0bc91289e03206e205b129b8a6eacae493cac73c19db35e74946a6db5816.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": "fede39a7bbfb37bbdfc4bff90b168bc3762769d624d0a34cafb88d4bff5c1910", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "6889bce3e8926da1ffd284fbb6ff2b7ccf1398de19c617112a09cc09dae34c21", "generation_seconds_batch": 17.580767795909196, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 4.399286270141602, "target_z": 1.3984376683930342, "ce": 5.065521717071533, "genuineness": 4.030367851257324, "identity": 0.2892473638057709, "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_DRAM_reg/fede39a7bbfb37bbdfc4bff90b168bc3762769d624d0a34cafb88d4bff5c1910.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-es-0", "key": "fef561ddad302493ca55cf4b4a37953824426ed68892c9310c668d4ed7c01101", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] La puerta sigue abierta.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "a26ac2b01118896e5aa9b3d880d36273bcf30cb99bd5962e67943e5b0834d612", "generation_seconds_batch": 16.19132740003988, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.996283769607544, "target_z": 0.1669921088212158, "ce": 6.038498878479004, "genuineness": 4.325342178344727, "identity": 0.3657665252685547, "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_DRAM_reg/fef561ddad302493ca55cf4b4a37953824426ed68892c9310c668d4ed7c01101.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": "ff486ad329bd0d2861915cd15adc2c42c992d94e3ba3abc0be2a76f16d56a10f", "bucket": "vn_S_DRAM_reg", "phase": "heldout", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] Someone moved the little blue chair.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "62b76b355303401e787de835bd06e9229cb5f5363ad77f6d4cbf8f41edb007a8", "generation_seconds_batch": 21.23578968003858, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.268873453140259, "target_z": 0.40624995422325283, "ce": 6.23706579208374, "genuineness": 4.268857479095459, "identity": 0.28646421432495117, "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": 0.03999999999999915, "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_DRAM_reg/ff486ad329bd0d2861915cd15adc2c42c992d94e3ba3abc0be2a76f16d56a10f.mp3" }, { "id": "vn_S_DRAM_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_DRAM_reg-nl-0", "key": "ff858b1b8e60806ba4f56a3d33fafa4fe13e0817f3d048d22985689e7d3ea773", "bucket": "vn_S_DRAM_reg", "phase": "grid", "candidate": "manual-21a285777a88", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Employ extreme dynamic shifts. Start almost whispered, suggesting secrecy or fear, then transition abruptly to a full, resonant declaration.\nSCRIPT:\n(Initiate at the very edge of audibility, focusing on breath control for intimate delivery.) [3.2 seconds duration] De deur staat nog open.\n(Execute a sudden, powerful surge in volume and projection, making the shift jarring and intentional.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Hold the peak energy level briefly, then allow a controlled, slow dissipation of power.) [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": "f28b9184bf20718beeebd93bae9b5e31f3f6daf953666cb9777699217af37e82", "generation_seconds_batch": 25.079145175986923, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.89638614654541, "target_z": 0.9570313816081482, "ce": 6.054384231567383, "genuineness": 4.300237655639648, "identity": 0.4087802767753601, "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_DRAM_reg/ff858b1b8e60806ba4f56a3d33fafa4fe13e0817f3d048d22985689e7d3ea773.mp3" }, { "key": "original-vn_S_DRAM_reg-de-0-seed1701-scale0.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "c4c9ba6c91f0f5aa4481f37bdf2ce9ee5755e39b365a2c8580ef26fa72dc0df8", "task_id": "vn_S_DRAM_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.9083876609802246, "target_z": 0.08984379822907292, "ce": 6.284721851348877, "genuineness": 4.444586753845215, "identity": 0.3443950414657593, "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_DRAM_reg/de/ae3165239878918dd9495837.mp3" }, { "key": "original-vn_S_DRAM_reg-de-0-seed1701-scale0.5", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "dc9d29e518afc354cc413b1d36745891804a19fc5f06e8266aa3d26700f05a24", "task_id": "vn_S_DRAM_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.949554204940796, "target_z": 0.12597656188691855, "ce": 6.395919322967529, "genuineness": 4.557555198669434, "identity": 0.6527838706970215, "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_DRAM_reg/de/78054d25281ac3b1c4ac5e21.mp3" }, { "key": "original-vn_S_DRAM_reg-de-0-seed1701-scale0.75", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "0fbd6c2bbcab1d0a78818df5c33125aac0b1dbab2cfa16484f8bbb5b5c4aebed", "task_id": "vn_S_DRAM_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.3787848949432373, "target_z": -0.37499997384185874, "ce": 6.205295085906982, "genuineness": 4.325342178344727, "identity": 0.7918428778648376, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf dem 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/audio/vn_S_DRAM_reg/de/b1e0cec5fcd4c06010845641.mp3" }, { "key": "original-vn_S_DRAM_reg-de-0-seed1701-scale1.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "a1ae9d79b16eda1d03fe07fb148d66375b2119e4f22a61f48d2b343ef2e025fd", "task_id": "vn_S_DRAM_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.407712697982788, "target_z": -0.34960941709825855, "ce": 6.252951145172119, "genuineness": 4.381826400756836, "identity": 0.5654391050338745, "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_DRAM_reg/de/36cad5e71f3bb0167d70e480.mp3" }, { "key": "original-vn_S_DRAM_reg-de-0-seed1702-scale0.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "a2ee26652bffb440252a4fecfb822ffff28d3a52295a5949f1c2c54f2c1ecb18", "task_id": "vn_S_DRAM_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.628009796142578, "target_z": -0.15624990190697036, "ce": 6.268836498260498, "genuineness": 4.406930446624756, "identity": 0.4337800145149231, "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_DRAM_reg/de/edd959e66c59ee57d5fc70d5.mp3" }, { "key": "original-vn_S_DRAM_reg-de-0-seed1702-scale0.5", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "65df9165a5869ddba0cd010d7030bf3f83f26bd9fa82e8cfb5e4ea52506530f0", "task_id": "vn_S_DRAM_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.7448337078094482, "target_z": -0.05371103457122726, "ce": 6.229123115539551, "genuineness": 4.432034492492676, "identity": 0.4515417218208313, "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_DRAM_reg/de/03a34a27b1f663a81ad71895.mp3" }, { "key": "original-vn_S_DRAM_reg-de-0-seed1702-scale0.75", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "f014576f90fda216856b3275050d3814b2cea003d823abac009e3545af858470", "task_id": "vn_S_DRAM_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.1028573513031006, "target_z": -0.6171875310627927, "ce": 6.308549880981445, "genuineness": 4.394378662109375, "identity": 0.49205613136291504, "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_DRAM_reg/de/7029065a6159be59e6f1ba32.mp3" }, { "key": "original-vn_S_DRAM_reg-de-0-seed1702-scale1.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "82a1707434b8f44a3e5410f493243214cf8a547a97e0e2a528c574d48222c00e", "task_id": "vn_S_DRAM_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.4166135787963867, "target_z": -0.34179692200291006, "ce": 6.24500846862793, "genuineness": 4.582659721374512, "identity": 0.6444147229194641, "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_DRAM_reg/de/d0dccc41dce645b45dcae39a.mp3" }, { "key": "original-vn_S_DRAM_reg-de-1-seed1701-scale0.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "6cc51797ea9d24803f626221238d04c8f7c170fc0877f6c2152b2dac5bdf20e2", "task_id": "vn_S_DRAM_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.4499919414520264, "target_z": -0.3125000130790706, "ce": 6.292664527893066, "genuineness": 4.268857479095459, "identity": 0.587099552154541, "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_DRAM_reg/de/1f00ffa837285ead93ff062a.mp3" }, { "key": "original-vn_S_DRAM_reg-de-1-seed1701-scale0.5", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "32423de141c8e7d141c190bbf2b7b35101f8a12ad2668577fb9054188aa9f079", "task_id": "vn_S_DRAM_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.3765597343444824, "target_z": -0.3769530452994134, "ce": 6.292664527893066, "genuineness": 4.168440818786621, "identity": 0.6240904331207275, "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_DRAM_reg/de/77ca883562191b531694bdc9.mp3" }, { "key": "original-vn_S_DRAM_reg-de-1-seed1701-scale0.75", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "55637370f05ea98b421b3104486fe97cef8b12cacf0e51a265812a2ffe61741f", "task_id": "vn_S_DRAM_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.4188389778137207, "target_z": -0.3398436412802255, "ce": 6.419747352600098, "genuineness": 4.21237325668335, "identity": 0.4979897141456604, "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/audio/vn_S_DRAM_reg/de/aab9231773de78669db905fa.mp3" }, { "key": "original-vn_S_DRAM_reg-de-1-seed1701-scale1.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "f2382d3186fbb28e850dc523e387bc324c42debd08065b74af07e79e07abeb12", "task_id": "vn_S_DRAM_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.6469240188598633, "target_z": -0.13964848062006094, "ce": 6.4515180587768555, "genuineness": 4.557555198669434, "identity": 0.5154018998146057, "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/audio/vn_S_DRAM_reg/de/daf47ba249c2d7e029939b1e.mp3" }, { "key": "original-vn_S_DRAM_reg-de-1-seed1702-scale0.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "1029d7cbd8250dabe912a7a293d825262e1294f528cd59e2a1caa39259e4de8a", "task_id": "vn_S_DRAM_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.4522171020507812, "target_z": -0.31054694162151597, "ce": 6.356205940246582, "genuineness": 4.15588903427124, "identity": 0.3368716835975647, "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_DRAM_reg/de/77fe7250827cd306e0048029.mp3" }, { "key": "original-vn_S_DRAM_reg-de-1-seed1702-scale0.5", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "a98592189662c05d684546966287bab496e8c1d1c3c214576405c94498df29d1", "task_id": "vn_S_DRAM_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.512298107147217, "target_z": -0.257812547411631, "ce": 6.197352409362793, "genuineness": 4.218649387359619, "identity": 0.5417424440383911, "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_DRAM_reg/de/c72cfe61f0af713e972c396d.mp3" }, { "key": "original-vn_S_DRAM_reg-de-1-seed1702-scale0.75", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "63059762b538fe7011b116275f006a42d8d30ef85a4cdb685f0787f0f02cd176", "task_id": "vn_S_DRAM_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.316478729248047, "target_z": -0.4296874395092984, "ce": 6.41180419921875, "genuineness": 4.256305694580078, "identity": 0.5871543288230896, "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_DRAM_reg/de/764eef2c10fc024edbfcac37.mp3" }, { "key": "original-vn_S_DRAM_reg-de-1-seed1702-scale1.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "85cdcfd566c5f1936f1da3c6c295f7289b5223eccf09e1a084371b7bbf7ea365", "task_id": "vn_S_DRAM_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.4277396202087402, "target_z": -0.3320313554500069, "ce": 6.348263263702393, "genuineness": 4.369274139404297, "identity": 0.487271785736084, "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_DRAM_reg/de/4ede3cb12149d66ef7c9815c.mp3" }, { "key": "original-vn_S_DRAM_reg-en-0-seed1701-scale0.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "5aa38e8ac27a30dbf7a0ca96467b70aa753b7a8cfa47ad779dd18640866da159", "task_id": "vn_S_DRAM_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.9840450286865234, "target_z": 0.15624990190697036, "ce": 6.23706579208374, "genuineness": 3.961331605911255, "identity": 0.3374236822128296, "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_DRAM_reg/en/3a94f443b5f0e8c26f72fa58.mp3" }, { "key": "original-vn_S_DRAM_reg-en-0-seed1701-scale0.5", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "023af6d91b6a21223e8b425bf7e6096520b44bd5e505d4289da5357479fcef9d", "task_id": "vn_S_DRAM_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.975144147872925, "target_z": 0.14843740681162182, "ce": 6.181467056274414, "genuineness": 3.948779582977295, "identity": 0.3924868702888489, "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_DRAM_reg/en/232c30b3075e9150202676c2.mp3" }, { "key": "original-vn_S_DRAM_reg-en-0-seed1701-scale0.75", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "bbf5c7f3a61568133f9a66484c67fa25558f05eab55609cf3f07408cacf13c58", "task_id": "vn_S_DRAM_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.895036220550537, "target_z": 0.07812495095348518, "ce": 6.205295085906982, "genuineness": 3.7416701316833496, "identity": 0.42365720868110657, "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_DRAM_reg/en/78c516f130a9f37897d21fc9.mp3" }, { "key": "original-vn_S_DRAM_reg-en-0-seed1701-scale1.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "aa8ddfdcd04ee6061bd203633088c1401fa5bc19ad0bb4876ae6aa1f85acb438", "task_id": "vn_S_DRAM_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 3.0530269145965576, "target_z": 0.21679679121220383, "ce": 6.141753673553467, "genuineness": 3.8922953605651855, "identity": 0.383301317691803, "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_DRAM_reg/en/3b61388dd69393e7ba26edc0.mp3" }, { "key": "original-vn_S_DRAM_reg-en-0-seed1702-scale0.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "63299ab063ab22cd6802ea150c9d0b276a63bb02fdb385b5c36cb24f7a055115", "task_id": "vn_S_DRAM_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.8229947090148926, "target_z": 0.014892562235972802, "ce": 6.149696350097656, "genuineness": 3.8169827461242676, "identity": 0.34863173961639404, "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_DRAM_reg/en/552edf66e935f2cf821fe980.mp3" }, { "key": "original-vn_S_DRAM_reg-en-0-seed1702-scale0.5", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "75a5f316771303fda66ebcd5c1f8250fa7781ee423eb00966beef08b8d3ee079", "task_id": "vn_S_DRAM_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.9617929458618164, "target_z": 0.13671876880116401, "ce": 6.125868320465088, "genuineness": 4.055472373962402, "identity": 0.361173152923584, "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_DRAM_reg/en/a0bb5f607e36c184f89d574a.mp3" }, { "key": "original-vn_S_DRAM_reg-en-0-seed1702-scale0.75", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "bed8bd1025e843c6976130f0741ea3137b41577905b1889d567cd563d6930044", "task_id": "vn_S_DRAM_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.658050060272217, "target_z": -0.1298829140671578, "ce": 6.260893821716309, "genuineness": 3.9801597595214844, "identity": 0.5618414878845215, "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_DRAM_reg/en/223d81f82e0f8016eefd4841.mp3" }, { "key": "original-vn_S_DRAM_reg-en-0-seed1702-scale1.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "7bd3cd041ced1d17e1a12130941752ca8caf97c7df2d6d255b50e6df37cfbe62", "task_id": "vn_S_DRAM_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.9851577281951904, "target_z": 0.15722654226831265, "ce": 6.38797664642334, "genuineness": 4.18099308013916, "identity": 0.2190553843975067, "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_DRAM_reg/en/69babb36bbe89c847dc686c2.mp3" }, { "key": "original-vn_S_DRAM_reg-en-1-seed1701-scale0.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "a8d6346c0385ec4cc41a8faa6a6c7402a7f105ae109327b9a28002711ac552d5", "task_id": "vn_S_DRAM_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.5011720657348633, "target_z": -0.2675781139645342, "ce": 6.340320587158203, "genuineness": 4.3881025314331055, "identity": 0.40914809703826904, "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_DRAM_reg/en/5137be46a95d96d7d9bcec66.mp3" }, { "key": "original-vn_S_DRAM_reg-en-1-seed1701-scale0.5", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "4a39a62ca7280ec5669d1d29b17ef9f8f4d46e75464e4fc332d8b30f19d29da2", "task_id": "vn_S_DRAM_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.287550926208496, "target_z": -0.4550779962528986, "ce": 6.205295085906982, "genuineness": 4.080576419830322, "identity": 0.5049508213996887, "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_DRAM_reg/en/61124ffedafbf50e7087c455.mp3" }, { "key": "original-vn_S_DRAM_reg-en-1-seed1701-scale0.75", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "57307cc6858bf0bd479e73405469fef60efff38f8813c3ad160acb91d9d048ac", "task_id": "vn_S_DRAM_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.3698840141296387, "target_z": -0.3828124689372073, "ce": 6.395919322967529, "genuineness": 4.425758361816406, "identity": 0.5733446478843689, "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_DRAM_reg/en/e5d3dbb5fb2ba899229fa225.mp3" }, { "key": "original-vn_S_DRAM_reg-en-1-seed1701-scale1.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "147b16955962535e5790c2e70d9d814e2b3c9e544cf97f4db6badc7dc7b310cf", "task_id": "vn_S_DRAM_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.5957438945770264, "target_z": -0.18457037973459742, "ce": 6.507116317749023, "genuineness": 4.406930446624756, "identity": 0.47887325286865234, "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_DRAM_reg/en/83258f865530215ffb1270a5.mp3" }, { "key": "original-vn_S_DRAM_reg-en-1-seed1702-scale0.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": 127, "seconds_generated": 10.16, "ended": true, "contract_sha256": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "b6890f3a3776bca4ab72723f62d526d4c62a52e0bccd6f0d60ec9e77e2eca7eb", "task_id": "vn_S_DRAM_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.65248703956604, "target_z": -0.13476569734360938, "ce": 6.324435234069824, "genuineness": 4.193545341491699, "identity": 0.2741979658603668, "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_DRAM_reg/en/b6f1f2168e5020026455e7c8.mp3" }, { "key": "original-vn_S_DRAM_reg-en-1-seed1702-scale0.5", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": 127, "seconds_generated": 10.16, "ended": true, "contract_sha256": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "9d9bed46a86e030dbadbf0954cadadac3a711bde0ec76c4f5ef37ff5cab13204", "task_id": "vn_S_DRAM_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.703667163848877, "target_z": -0.08984379822907292, "ce": 6.395919322967529, "genuineness": 4.482243061065674, "identity": 0.1562424749135971, "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_DRAM_reg/en/0465f3c7ed5266ac59c5fabd.mp3" }, { "key": "original-vn_S_DRAM_reg-en-1-seed1702-scale0.75", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "e3866e8e6018e48f6de019baf0648ec2ba097a2a4accdeb60955d71cdc576f8c", "task_id": "vn_S_DRAM_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.4032623767852783, "target_z": -0.35351556001336787, "ce": 6.173524379730225, "genuineness": 3.936227321624756, "identity": 0.33247172832489014, "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_DRAM_reg/en/bcd2fbff63f29dc6ae120eb7.mp3" }, { "key": "original-vn_S_DRAM_reg-en-1-seed1702-scale1.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": 127, "seconds_generated": 10.16, "ended": true, "contract_sha256": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "2ec1c124485ac801d9c9c2a422aef5644cbb5408d4ef64e5a8b2679ba47b13aa", "task_id": "vn_S_DRAM_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.429965019226074, "target_z": -0.3300780747273223, "ce": 6.324435234069824, "genuineness": 4.256305694580078, "identity": 0.46709489822387695, "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_DRAM_reg/en/6e3d44d2e3ef1ec58023777b.mp3" }, { "key": "original-vn_S_DRAM_reg-es-0-seed1701-scale0.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "1b60cf123f76d948227125c1813512cc570662714a54a189e5cc961826a4951d", "task_id": "vn_S_DRAM_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 3.1509368419647217, "target_z": 0.30273444652616743, "ce": 6.022613525390625, "genuineness": 4.174716949462891, "identity": 0.42868533730506897, "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_DRAM_reg/es/039289feb258e0054e9c117a.mp3" }, { "key": "original-vn_S_DRAM_reg-es-0-seed1701-scale0.5", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "8591a068bb742334e3b19d1866378b794495e612b19d7a87be9bce994a80e6de", "task_id": "vn_S_DRAM_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.946216344833374, "target_z": 0.12304685006802163, "ce": 6.086154937744141, "genuineness": 4.080576419830322, "identity": 0.738068699836731, "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_DRAM_reg/es/4bf52823ef218fdebd8d743d.mp3" }, { "key": "original-vn_S_DRAM_reg-es-0-seed1701-scale0.75", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "61bddb6b437464e82b2ae73784ddad48425c31567054ec2e56a5416f4d741fd0", "task_id": "vn_S_DRAM_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.705892562866211, "target_z": -0.08789051750638834, "ce": 5.792276382446289, "genuineness": 3.71028995513916, "identity": 0.4844037890434265, "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_DRAM_reg/es/3558423b0f35b294b0dc6731.mp3" }, { "key": "original-vn_S_DRAM_reg-es-0-seed1701-scale1.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "04df105887d2e9323cfc240661739ad46b1a865000c7ffe92408a54922e12de6", "task_id": "vn_S_DRAM_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.632460117340088, "target_z": -0.15234375899186106, "ce": 6.133810997009277, "genuineness": 3.9425034523010254, "identity": 0.5866736173629761, "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_DRAM_reg/es/f22fc73143714474c31e04a3.mp3" }, { "key": "original-vn_S_DRAM_reg-es-0-seed1702-scale0.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "e2350911f543088df537c2b9657e3d2bf2283f2fde4f5380140792ed877698f5", "task_id": "vn_S_DRAM_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.947885274887085, "target_z": 0.12451170597747009, "ce": 5.974957466125488, "genuineness": 4.005263805389404, "identity": 0.39446863532066345, "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_DRAM_reg/es/a4c8c9b28427d5195f96e513.mp3" }, { "key": "original-vn_S_DRAM_reg-es-0-seed1702-scale0.5", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": 127, "seconds_generated": 10.16, "ended": true, "contract_sha256": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "920a47e738345e8d3bc2b7e7738bd3143d5ad8bbd9e2394dfad06bd42e81030a", "task_id": "vn_S_DRAM_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 3.242170810699463, "target_z": 0.3828124689372073, "ce": 6.046441555023193, "genuineness": 4.18726921081543, "identity": 0.5948216915130615, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/es/e9e2d7a0f12742e8440886a2.mp3" }, { "key": "original-vn_S_DRAM_reg-es-0-seed1702-scale0.75", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "a72b8f94706f281b51ff6b1b1ac1c3a9cd667a85322cb50153872242a9c51f8a", "task_id": "vn_S_DRAM_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.8071401119232178, "target_z": 0.000976640361342285, "ce": 6.054384231567383, "genuineness": 3.873466968536377, "identity": 0.4038418233394623, "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_DRAM_reg/es/bfc2223afb9b96cc6a06dba3.mp3" }, { "key": "original-vn_S_DRAM_reg-es-0-seed1702-scale1.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": 127, "seconds_generated": 10.16, "ended": true, "contract_sha256": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "a5573f15606f02cad913edd76ee934212c6e63f9e2c0e140c365b99231fcbc7f", "task_id": "vn_S_DRAM_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.7448337078094482, "target_z": -0.05371103457122726, "ce": 5.974957466125488, "genuineness": 4.105680465698242, "identity": 0.573773980140686, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "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_DRAM_reg/es/96dd9cd93c8844a0572b264c.mp3" }, { "key": "original-vn_S_DRAM_reg-es-1-seed1701-scale0.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "9a6fc02359ff28e6d6bbfe7535d0e8aa4f29ffde0509391c0d3b1b8ce8b629b6", "task_id": "vn_S_DRAM_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.6435861587524414, "target_z": -0.1425781924389579, "ce": 6.149696350097656, "genuineness": 4.419482707977295, "identity": 0.7322819232940674, "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_DRAM_reg/es/c7284f383f64724d69c421b0.mp3" }, { "key": "original-vn_S_DRAM_reg-es-1-seed1701-scale0.5", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "e0a6cfb204b5b6327e52f1de68d3612da258f421cee20bf6ce2bf70310ca5271", "task_id": "vn_S_DRAM_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.4277396202087402, "target_z": -0.3320313554500069, "ce": 6.046441555023193, "genuineness": 4.381826400756836, "identity": 0.6569411158561707, "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_DRAM_reg/es/38b40fb37e6f30450530dc86.mp3" }, { "key": "original-vn_S_DRAM_reg-es-1-seed1701-scale0.75", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "530fbfd4312c93615063c51fe5c8bedc4f032d67b87286d27c10863eece94662", "task_id": "vn_S_DRAM_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.0271997451782227, "target_z": -0.6835938440058201, "ce": 6.0305562019348145, "genuineness": 4.607763767242432, "identity": 0.7792811989784241, "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_DRAM_reg/es/33e5c9b2490abe104d843059.mp3" }, { "key": "original-vn_S_DRAM_reg-es-1-seed1701-scale1.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "77de66cda8f0e4d04ed72da90ba29ad86cfd49e655f791cf661be9f32a0d434a", "task_id": "vn_S_DRAM_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.5434513092041016, "target_z": -0.23046870994534624, "ce": 6.0305562019348145, "genuineness": 4.325342178344727, "identity": 0.6923272609710693, "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_DRAM_reg/es/4075a0af20bd15d8eea53d2e.mp3" }, { "key": "original-vn_S_DRAM_reg-es-1-seed1702-scale0.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "3b632a98ffeed8e706cfc1c9d6e494c093e597470bd226b0fa13af274a6c81ca", "task_id": "vn_S_DRAM_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.73148250579834, "target_z": -0.06542967258168507, "ce": 6.173524379730225, "genuineness": 4.193545341491699, "identity": 0.6161152124404907, "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_DRAM_reg/es/6e9e28c72e7f6d972fb6d814.mp3" }, { "key": "original-vn_S_DRAM_reg-es-1-seed1702-scale0.5", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "0990d5a1233781c9e0e322270a4dc374d811e5796d092ca77f6ea1ee7e4a58ba", "task_id": "vn_S_DRAM_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.6880905628204346, "target_z": -0.1035157169622153, "ce": 6.165581703186035, "genuineness": 4.20609712600708, "identity": 0.39224135875701904, "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_DRAM_reg/es/ec0751cb4536fd129033e373.mp3" }, { "key": "original-vn_S_DRAM_reg-es-1-seed1702-scale0.75", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "8070a7facd2205b52ce7e1413d93d666d7a983307d9fe9228936e27da6ba3e7c", "task_id": "vn_S_DRAM_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.4232892990112305, "target_z": -0.33593749836511616, "ce": 6.221180438995361, "genuineness": 4.168440818786621, "identity": 0.5236681699752808, "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_DRAM_reg/es/a0248b6ae4715e0d3db5b20b.mp3" }, { "key": "original-vn_S_DRAM_reg-es-1-seed1702-scale1.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "0e01e7395d80dfb11b953a517ec8f6ee7137b1efdb0d816b1c57b139ac40b463", "task_id": "vn_S_DRAM_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.409937858581543, "target_z": -0.3476563456407039, "ce": 6.046441555023193, "genuineness": 4.381826400756836, "identity": 0.6967529058456421, "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_DRAM_reg/es/20da2787dcbaae9f51372a53.mp3" }, { "key": "original-vn_S_DRAM_reg-fr-0-seed1701-scale0.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "1b66bcca1ca6c8b7e298a5d6ab1bb4cfe9855953280483b55937f64b918908cf", "task_id": "vn_S_DRAM_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.3431813716888428, "target_z": -0.40624995422325283, "ce": 6.205295085906982, "genuineness": 4.632867813110352, "identity": 0.7958292961120605, "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_DRAM_reg/fr/654f3abc3125b0eedf5c4f32.mp3" }, { "key": "original-vn_S_DRAM_reg-fr-0-seed1701-scale0.5", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "2aa9c40049df10839a4baf7c8495db4048e725d776649e42996ce57f2c787f34", "task_id": "vn_S_DRAM_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.579054832458496, "target_z": -0.19921872956395215, "ce": 6.1020402908325195, "genuineness": 4.256305694580078, "identity": 0.8360067009925842, "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_DRAM_reg/fr/581d27a0e6efb22d6ed094fb.mp3" }, { "key": "original-vn_S_DRAM_reg-fr-0-seed1701-scale0.75", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "c75c3183ce195d22c9dd6145fddc59790db4c2484ec2cb1b4d099ed74b1916bb", "task_id": "vn_S_DRAM_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.9139506816864014, "target_z": 0.0947265815055245, "ce": 6.165581703186035, "genuineness": 3.9425034523010254, "identity": 0.7637378573417664, "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/audio/vn_S_DRAM_reg/fr/2a8066a0d68d90d8746f5cbc.mp3" }, { "key": "original-vn_S_DRAM_reg-fr-0-seed1701-scale1.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "459c1845270e9b8fbaefea0d021a133e4adc7bed731924797848bdeb751d57bb", "task_id": "vn_S_DRAM_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.850531816482544, "target_z": 0.03906247547674259, "ce": 5.935244083404541, "genuineness": 4.080576419830322, "identity": 0.7025177478790283, "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_DRAM_reg/fr/aebd9a9da8c9caa11f64fac9.mp3" }, { "key": "original-vn_S_DRAM_reg-fr-0-seed1702-scale0.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "ebc4a5d086db777889edbce0eaf824aee8d1507e2dfa8cf101d40e7556092348", "task_id": "vn_S_DRAM_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.852200746536255, "target_z": 0.04052733138619105, "ce": 6.109982967376709, "genuineness": 4.231201171875, "identity": 0.6346712112426758, "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_DRAM_reg/fr/adc7a3c0dff3056857d150d2.mp3" }, { "key": "original-vn_S_DRAM_reg-fr-0-seed1702-scale0.5", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "1464bb6fa39c63c32d97e619b020ec6ba7e2680ce59644a0f82b7b010b012749", "task_id": "vn_S_DRAM_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.8513662815093994, "target_z": 0.03979490343146682, "ce": 6.213237762451172, "genuineness": 4.444586753845215, "identity": 0.701950192451477, "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_DRAM_reg/fr/5513bfc9ec45f44551d357d1.mp3" }, { "key": "original-vn_S_DRAM_reg-fr-0-seed1702-scale0.75", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "12c69f821009e3abf261e6f055cd3fec20b7869bc3be0dd2e2fb44066bf94ced", "task_id": "vn_S_DRAM_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.7192437648773193, "target_z": -0.07617187949593053, "ce": 6.173524379730225, "genuineness": 4.369274139404297, "identity": 0.5007668733596802, "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_DRAM_reg/fr/c24686f76136f353902b99ea.mp3" }, { "key": "original-vn_S_DRAM_reg-fr-0-seed1702-scale1.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "281f8bd86a620d9624780a55575782df09198049ea76ddbdf873930b20c68741", "task_id": "vn_S_DRAM_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 3.0018467903137207, "target_z": 0.17187489209766738, "ce": 5.998785495758057, "genuineness": 4.394378662109375, "identity": 0.6903053522109985, "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_DRAM_reg/fr/41329270e21027c7a7d41ba4.mp3" }, { "key": "original-vn_S_DRAM_reg-fr-1-seed1701-scale0.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "f3abd2b5f935946e2e3d4da8f29a832d059475a836e4c19c4d252f29e58ba865", "task_id": "vn_S_DRAM_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.868333578109741, "target_z": 0.05468746566743962, "ce": 6.197352409362793, "genuineness": 4.406930446624756, "identity": 0.3949604332447052, "hypothesis": "J'ai trouvé la petite Pacilboite. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble what it contient.", "wer": 0.15789473684210525, "duration_error_s": 0.040000000000000924, "model_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_DRAM_reg/fr/fe355f8305ace8205b6526e0.mp3" }, { "key": "original-vn_S_DRAM_reg-fr-1-seed1701-scale0.5", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "d45b6c8c74e06cd66c987704f49dbd640403c5539c15ba5b383f385e00e19e07", "task_id": "vn_S_DRAM_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.9528920650482178, "target_z": 0.1289062737058155, "ce": 6.046441555023193, "genuineness": 4.300237655639648, "identity": 0.7645845413208008, "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_DRAM_reg/fr/081c329d3c13c94cd27d9544.mp3" }, { "key": "original-vn_S_DRAM_reg-fr-1-seed1701-scale0.75", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "67778f1543ec0c53136077c02ac38d4509d6356f53be4c0980feb679b110cd6a", "task_id": "vn_S_DRAM_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.481145143508911, "target_z": -0.28515617561278583, "ce": 5.974957466125488, "genuineness": 4.482243061065674, "identity": 0.4950774312019348, "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_DRAM_reg/fr/5a78cd9c7755d58b6a50eef3.mp3" }, { "key": "original-vn_S_DRAM_reg-fr-1-seed1701-scale1.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "d0ac1409bf78a58028c2a5a8b39b30cd41b6a0ae6359bd85ea4b4067c2f5d4f5", "task_id": "vn_S_DRAM_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.704779863357544, "target_z": -0.08886715786773063, "ce": 6.022613525390625, "genuineness": 4.344170093536377, "identity": 0.7311161756515503, "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_DRAM_reg/fr/25ae0c80f63dfcd1253abfbf.mp3" }, { "key": "original-vn_S_DRAM_reg-fr-1-seed1702-scale0.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "23a5a1f822ddd7f115ae5d4569bcfd1f44fb403a35c11905cd5b52b0f24a94e5", "task_id": "vn_S_DRAM_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.625784397125244, "target_z": -0.1582031826296549, "ce": 6.221180438995361, "genuineness": 4.507347106933594, "identity": 0.8053501844406128, "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_DRAM_reg/fr/dbc22dd837df8eb305e66422.mp3" }, { "key": "original-vn_S_DRAM_reg-fr-1-seed1702-scale0.5", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "806368a56cf853363b1e0f4139c018d2aa70b2ef8554547c7d050ab026a4aa04", "task_id": "vn_S_DRAM_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.8099215030670166, "target_z": 0.003417927367003116, "ce": 6.300607204437256, "genuineness": 4.432034492492676, "identity": 0.7234939336776733, "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_DRAM_reg/fr/e37def329f863642e788cc73.mp3" }, { "key": "original-vn_S_DRAM_reg-fr-1-seed1702-scale0.75", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "1a0b5b9440fa0842c739752d8da676feb75c22162a41d91dc281a8366f61c3b4", "task_id": "vn_S_DRAM_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.650261878967285, "target_z": -0.13671876880116401, "ce": 6.332377910614014, "genuineness": 4.68307638168335, "identity": 0.7744957208633423, "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_DRAM_reg/fr/7a5ae1c50efed3ca958f25d3.mp3" }, { "key": "original-vn_S_DRAM_reg-fr-1-seed1702-scale1.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "f574763a5a15fc39cad5e77a6dedc290c3a3bd21a754e01f81b433bd6e3e2c8f", "task_id": "vn_S_DRAM_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.5634782314300537, "target_z": -0.21289064829709453, "ce": 6.141753673553467, "genuineness": 4.149612903594971, "identity": 0.552850604057312, "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_DRAM_reg/fr/0182867440a2d83d9d6d256e.mp3" }, { "key": "original-vn_S_DRAM_reg-nl-0-seed1701-scale0.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "e3d35928d01203ea3f2c1a7decaaaa9dcadeea944f816a7d31b08afea607207f", "task_id": "vn_S_DRAM_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 3.998746156692505, "target_z": 1.046874970572091, "ce": 6.141753673553467, "genuineness": 4.093128204345703, "identity": 0.6036685109138489, "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_DRAM_reg/nl/ca8f9ddf1b19fc272a49e27d.mp3" }, { "key": "original-vn_S_DRAM_reg-nl-0-seed1701-scale0.5", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "ae64806145d5b4d104931d4eec915163fb5f9fd057ef84f87ab4d7f17da2abae", "task_id": "vn_S_DRAM_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.949554204940796, "target_z": 0.12597656188691855, "ce": 6.046441555023193, "genuineness": 4.193545341491699, "identity": 0.7222456932067871, "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": 0.03999999999999915, "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_DRAM_reg/nl/191c1d0d8bde7417e385a310.mp3" }, { "key": "original-vn_S_DRAM_reg-nl-0-seed1701-scale0.75", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "53193abe6b26e93e6c62cc66fa435ce5b3ad65ee3cd5f4a805c9dae8a53cd233", "task_id": "vn_S_DRAM_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.8108949661254883, "target_z": 0.004272356892471415, "ce": 6.109982967376709, "genuineness": 4.3504462242126465, "identity": 0.607108473777771, "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_DRAM_reg/nl/9f366e7a5fd91fc1e1b1d29a.mp3" }, { "key": "original-vn_S_DRAM_reg-nl-0-seed1701-scale1.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "8d975aef0363d08750fc658b343ad783788420a86bff3e83f13ea0fce1707f13", "task_id": "vn_S_DRAM_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 3.166513442993164, "target_z": 0.3164063652593098, "ce": 6.086154937744141, "genuineness": 4.2374773025512695, "identity": 0.738023042678833, "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_DRAM_reg/nl/ebd720eb767b545d9eaddc54.mp3" }, { "key": "original-vn_S_DRAM_reg-nl-0-seed1702-scale0.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "ca166c633e7037a93e0d5bdbe4b5f46e2940b79e649165fffaf05dbc2514cbf1", "task_id": "vn_S_DRAM_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.483370304107666, "target_z": -0.2832031041552312, "ce": 6.006728172302246, "genuineness": 4.18099308013916, "identity": 0.7997298240661621, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Wanneer je klaar bent.", "wer": 0.10526315789473684, "duration_error_s": 0.040000000000000924, "model_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_DRAM_reg/nl/c95656f086478b69debc8752.mp3" }, { "key": "original-vn_S_DRAM_reg-nl-0-seed1702-scale0.5", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "bae8fd311278b1ed760054384140f079b96081cabf68a8b8bbc86a338e2b30d3", "task_id": "vn_S_DRAM_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 3.115333318710327, "target_z": 0.2714844661447734, "ce": 6.133810997009277, "genuineness": 4.124508857727051, "identity": 0.7765477299690247, "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_DRAM_reg/nl/f962e7b7afb8e8f7f3195c1c.mp3" }, { "key": "original-vn_S_DRAM_reg-nl-0-seed1702-scale0.75", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "14ccb1aad3549c289399563cd0d0030454e161091737c1564a2f448fd8cf7c61", "task_id": "vn_S_DRAM_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 3.1820898056030273, "target_z": 0.3300780747273223, "ce": 6.181467056274414, "genuineness": 4.256305694580078, "identity": 0.7898513674736023, "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_DRAM_reg/nl/07db57eddd4abe3ed5fe234e.mp3" }, { "key": "original-vn_S_DRAM_reg-nl-0-seed1702-scale1.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "907dfd138b36761056e94a4a517298fb27b7f65294ef8c85d82b8460142e3a89", "task_id": "vn_S_DRAM_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 3.3267292976379395, "target_z": 0.45703127697558316, "ce": 6.141753673553467, "genuineness": 4.068024158477783, "identity": 0.7088188529014587, "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_DRAM_reg/nl/1640e53d6bb5f413e02a27a7.mp3" }, { "key": "original-vn_S_DRAM_reg-nl-1-seed1701-scale0.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "a9189b66b8c68569e8d70661700d4c3ac4fd592959f1c5b53252cc7224cd245a", "task_id": "vn_S_DRAM_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 3.162062883377075, "target_z": 0.3125000130790706, "ce": 6.308549880981445, "genuineness": 4.469690799713135, "identity": 0.6141202449798584, "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_DRAM_reg/nl/6691c86749c3cbffa92a8e96.mp3" }, { "key": "original-vn_S_DRAM_reg-nl-1-seed1701-scale0.5", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "9feee8154d93981db4d597eeff23bad306e1c27bcd8f98add3af7cf337c7638f", "task_id": "vn_S_DRAM_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 3.4023866653442383, "target_z": 0.5234373806534806, "ce": 6.284721851348877, "genuineness": 4.074300289154053, "identity": 0.6892604827880859, "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/audio/vn_S_DRAM_reg/nl/22622c622eec21bd2a125419.mp3" }, { "key": "original-vn_S_DRAM_reg-nl-1-seed1701-scale0.75", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "0a961a5f0ae9e1ec2ed6f1cd416920cbd4ab73c0ed1a93207ac13134f0471230", "task_id": "vn_S_DRAM_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.5879557132720947, "target_z": -0.19140623446860364, "ce": 6.1020402908325195, "genuineness": 4.243753433227539, "identity": 0.811199963092804, "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/audio/vn_S_DRAM_reg/nl/bc826c1fd273281caec00f59.mp3" }, { "key": "original-vn_S_DRAM_reg-nl-1-seed1701-scale1.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "f2ffd1699db6c5402ab0ff96a2b5ddf6d79c0c3a89b78dced04f52ee91b73eae", "task_id": "vn_S_DRAM_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 3.286675214767456, "target_z": 0.42187494441394985, "ce": 6.205295085906982, "genuineness": 4.287685871124268, "identity": 0.6544933319091797, "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_DRAM_reg/nl/7b6ba24368b8f00d344cb163.mp3" }, { "key": "original-vn_S_DRAM_reg-nl-1-seed1702-scale0.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "00743a887d80f5688c83c266843c03f0754637be8e7fc95034b2878426ea7515", "task_id": "vn_S_DRAM_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.831895589828491, "target_z": 0.02270505733132132, "ce": 6.221180438995361, "genuineness": 4.193545341491699, "identity": 0.6061031818389893, "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_DRAM_reg/nl/0da0f6b842ffbd2ca2617288.mp3" }, { "key": "original-vn_S_DRAM_reg-nl-1-seed1702-scale0.5", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "c9c0d24870bbd415201d17e2d7fcbc8b3e312ef43895b743ae466f9ac1b04e20", "task_id": "vn_S_DRAM_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.312028169631958, "target_z": -0.4335937916895376, "ce": 6.117925643920898, "genuineness": 4.306513786315918, "identity": 0.6626036763191223, "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_DRAM_reg/nl/01400a0224b79f61edb87504.mp3" }, { "key": "original-vn_S_DRAM_reg-nl-1-seed1702-scale0.75", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "46d2b0a902e2f03819c3f0c7200bad4715d95bfdd203741035317b316a26ee3a", "task_id": "vn_S_DRAM_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 3.0619277954101562, "target_z": 0.22460928630755236, "ce": 6.1020402908325195, "genuineness": 4.118232727050781, "identity": 0.590269148349762, "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_DRAM_reg/nl/8ac6bbca5d8b915545d8a778.mp3" }, { "key": "original-vn_S_DRAM_reg-nl-1-seed1702-scale1.0", "bucket": "vn_S_DRAM_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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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 Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Dramatic 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": "f83a7b6631e525260f619513535e905e45abd41c2d8fa7d2cc63332eb3e443dd", "mp3_sha256": "fb77a41228d75153aa31586a1323da793710a82e8ae52a1799c4ebd0604b5718", "task_id": "vn_S_DRAM_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Dramatic Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Dramatic 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 Dramatic 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 Dramatic 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": 2.5523521900177, "target_z": -0.2226562148499977, "ce": 6.213237762451172, "genuineness": 4.218649387359619, "identity": 0.6605888605117798, "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_DRAM_reg/nl/5c24c864b6b194fc11429fc3.mp3" } ], "source_standard_audio_unchanged": true }